{"Code":200,"Data":{"AigcAttributes":"{}","AigcIsTop":0,"AigcType":"","AlreadyStar":false,"ApplyMeta":"{}","ApprovalMode":1,"ApprovalNotifyEmail":"","Architectures":["Qwen3_5ForConditionalGeneration"],"Avatar":"https://resources.modelscope.cn/avatar/a70ddc46-13cb-40c1-8885-7fbb3875e38d.png","Backbone":[],"BackendSupport":{"architectures":null,"backend_info":{"deploy_task":null,"lmdeploy":null,"lmdeploy_turbomind":null,"ollama":null,"sglang":null,"vllm":null},"model_id":"Qwen/Qwen3.8-27B"},"BaseModel":[],"BaseModelRelation":"","CardReady":0,"CardUnreadyReason":"","CertificationCreateBy":"","CertificationCreatedTime":-62135596800,"ChineseName":"千问3.8-27B","CoverHeight":0,"CoverImages":[],"CoverWidth":0,"CreatedBy":"yingda","CreatedTime":1786526721,"DashSdkParameter":"","Datasets":{},"DemoAvailable":0,"DemoUnavailableReason":"","Description":"","Domain":[],"Downloads":1742,"ExampleCodeAvailable":0,"ExampleCodeUnavailableReason":"","ForbiddenVisibilityUpdate":false,"Frameworks":[],"FromSite":"maas","Id":1549152,"Integrating":0,"IntegrationFailureLog":"","IntegrationFailureReason":"","IsAccessible":1,"IsCertification":4,"IsHot":0,"IsNewModel":true,"IsOnline":1,"IsPreTrain":0,"IsPublished":1,"IsTop":99,"Language":[],"LastUpdatedTime":1786731897,"Libraries":["safetensors"],"License":"apache-2.0","LicenseLink":"","LicenseName":"","Meta":"","Metrics":[],"ModelDetail":{},"ModelInfos":{"safetensor":{"chat_template":"{%- set image_count = namespace(value=0) %}\n{%- set video_count = namespace(value=0) %}\n{%- macro render_content(content, do_vision_count, is_system_content=false) %}\n    {%- if content is string %}\n        {{- content }}\n    {%- elif content is iterable and content is not mapping %}\n        {%- for item in content %}\n            {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}\n                {%- if is_system_content %}\n                    {{- raise_exception('System message cannot contain images.') }}\n                {%- endif %}\n                {%- if do_vision_count %}\n                    {%- set image_count.value = image_count.value + 1 %}\n                {%- endif %}\n                {%- if add_vision_id %}\n                    {{- 'Picture ' ~ image_count.value ~ ': ' }}\n                {%- endif %}\n                {{- '\u003c|vision_start|\u003e\u003c|image_pad|\u003e\u003c|vision_end|\u003e' }}\n            {%- elif 'video' in item or item.type == 'video' %}\n                {%- if is_system_content %}\n                    {{- raise_exception('System message cannot contain videos.') }}\n                {%- endif %}\n                {%- if do_vision_count %}\n                    {%- set video_count.value = video_count.value + 1 %}\n                {%- endif %}\n                {%- if add_vision_id %}\n                    {{- 'Video ' ~ video_count.value ~ ': ' }}\n                {%- endif %}\n                {{- '\u003c|vision_start|\u003e\u003c|video_pad|\u003e\u003c|vision_end|\u003e' }}\n            {%- elif 'text' in item %}\n                {{- item.text }}\n            {%- else %}\n                {{- raise_exception('Unexpected item type in content.') }}\n            {%- endif %}\n        {%- endfor %}\n    {%- elif content is none or content is undefined %}\n        {{- '' }}\n    {%- else %}\n        {{- raise_exception('Unexpected content type.') }}\n    {%- endif %}\n{%- endmacro %}\n{%- if not messages %}\n    {{- raise_exception('No messages provided.') }}\n{%- endif %}\n{%- set reasoning_instructions = '' %}\n{%- if enable_thinking is undefined or enable_thinking is true %}\n    {%- set resolved_reasoning_effort = reasoning_effort|default('xhigh') %}\n    {%- if resolved_reasoning_effort not in ('xhigh', 'medium', 'low') %}\n        {{- raise_exception('Unexpected reasoning effort ' ~ reasoning_effort ~ '. Supported types are xhigh (default), medium, and low.') }}\n    {%- endif %}\n    {%- if resolved_reasoning_effort == 'xhigh' %}\n        {%- set reasoning_instructions = 'Reasoning effort is set to xhigh. Please think carefully through the task, validate key assumptions, consider plausible alternatives, and prioritize correctness, consistency, and clarity in the final answer.' %}\n    {%- elif resolved_reasoning_effort == 'low' %}\n        {%- set reasoning_instructions = 'Reasoning effort is set to low. Keep your thinking brief and focused, moving directly to the conclusion without unnecessary elaboration.' %}\n    {%- endif %}\n{%- endif %}\n{%- if tools and tools is iterable and tools is not mapping %}\n    {{- '\u003c|im_start|\u003esystem\\n' }}\n    {%- if reasoning_instructions %}\n        {{- reasoning_instructions + '\\n\\n' }}\n    {%- endif %}\n    {{- \"# Tools\\n\\nYou have access to the following functions:\\n\\n\u003ctools\u003e\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n\u003c/tools\u003e\" }}\n    {{- '\\n\\nIf you choose to call a function ONLY reply in the following format with NO suffix:\\n\\n\u003ctool_call\u003e\\n\u003cfunction=example_function_name\u003e\\n\u003cparameter=example_parameter_1\u003e\\nvalue_1\\n\u003c/parameter\u003e\\n\u003cparameter=example_parameter_2\u003e\\nThis is the value for the second parameter\\nthat can span\\nmultiple lines\\n\u003c/parameter\u003e\\n\u003c/function\u003e\\n\u003c/tool_call\u003e\\n\\n\u003cIMPORTANT\u003e\\nReminder:\\n- Function calls MUST follow the specified format: an inner \u003cfunction=...\u003e\u003c/function\u003e block must be nested within \u003ctool_call\u003e\u003c/tool_call\u003e XML tags\\n- Required parameters MUST be specified\\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\\n\u003c/IMPORTANT\u003e' }}\n    {%- if messages[0].role == 'system' %}\n        {%- set content = render_content(messages[0].content, false, true)|trim %}\n        {%- if content %}\n            {{- '\\n\\n' + content }}\n        {%- endif %}\n    {%- endif %}\n    {{- '\u003c|im_end|\u003e\\n' }}\n{%- else %}\n    {%- if messages[0].role == 'system' %}\n        {%- set content = render_content(messages[0].content, false, true)|trim %}\n        {%- if content %}\n            {{- '\u003c|im_start|\u003esystem\\n' + (reasoning_instructions + '\\n\\n' if reasoning_instructions else '')  + content + '\u003c|im_end|\u003e\\n' }}\n        {%- elif reasoning_instructions %}\n            {{- '\u003c|im_start|\u003esystem\\n' + reasoning_instructions + '\u003c|im_end|\u003e\\n' }}\n        {%- endif %}\n    {%- elif reasoning_instructions %}\n        {{- '\u003c|im_start|\u003esystem\\n' + reasoning_instructions + '\u003c|im_end|\u003e\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- if ns.multi_step_tool and message.role == \"user\" %}\n        {%- set content = render_content(message.content, false)|trim %}\n        {%- if not(content.startswith('\u003ctool_response\u003e') and content.endswith('\u003c/tool_response\u003e')) %}\n            {%- set ns.multi_step_tool = false %}\n            {%- set ns.last_query_index = index %}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if ns.multi_step_tool %}\n    {{- raise_exception('No user query found in messages.') }}\n{%- endif %}\n{%- for message in messages %}\n    {%- set content = render_content(message.content, true)|trim %}\n    {%- if message.role == \"system\" %}\n        {%- if not loop.first %}\n            {{- raise_exception('System message must be at the beginning.') }}\n        {%- endif %}\n    {%- elif message.role == \"user\" %}\n        {{- '\u003c|im_start|\u003e' + message.role + '\\n' + content + '\u003c|im_end|\u003e' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is string %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- endif %}\n        {%- set reasoning_content = reasoning_content|trim %}\n        {%- if preserve_thinking is undefined or preserve_thinking is true or loop.index0 \u003e ns.last_query_index %}\n            {{- '\u003c|im_start|\u003e' + message.role + '\\n\u003cthink\u003e\\n' + reasoning_content + '\\n\u003c/think\u003e\\n\\n' + content }}\n        {%- else %}\n            {{- '\u003c|im_start|\u003e' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if tool_call.function is defined %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {%- if loop.first %}\n                    {%- if content|trim %}\n                        {{- '\\n\\n\u003ctool_call\u003e\\n\u003cfunction=' + tool_call.name + '\u003e\\n' }}\n                    {%- else %}\n                        {{- '\u003ctool_call\u003e\\n\u003cfunction=' + tool_call.name + '\u003e\\n' }}\n                    {%- endif %}\n                {%- else %}\n                    {{- '\\n\u003ctool_call\u003e\\n\u003cfunction=' + tool_call.name + '\u003e\\n' }}\n                {%- endif %}\n                {%- if tool_call.arguments is defined and tool_call.arguments != '' %}\n                    {%- for args_name, args_value in tool_call.arguments|items %}\n                        {{- '\u003cparameter=' + args_name + '\u003e\\n' }}\n                        {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}\n                        {{- args_value }}\n                        {{- '\\n\u003c/parameter\u003e\\n' }}\n                    {%- endfor %}\n                {%- endif %}\n                {{- '\u003c/function\u003e\\n\u003c/tool_call\u003e' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '\u003c|im_end|\u003e\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.previtem and loop.previtem.role != \"tool\" %}\n            {{- '\u003c|im_start|\u003euser' }}\n        {%- endif %}\n        {{- '\\n\u003ctool_response\u003e\\n' }}\n        {{- content }}\n        {{- '\\n\u003c/tool_response\u003e' }}\n        {%- if not loop.last and loop.nextitem.role != \"tool\" %}\n            {{- '\u003c|im_end|\u003e\\n' }}\n        {%- elif loop.last %}\n            {{- '\u003c|im_end|\u003e\\n' }}\n        {%- endif %}\n    {%- else %}\n        {{- raise_exception('Unexpected message role.') }}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '\u003c|im_start|\u003eassistant\\n' }}\n    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '\u003cthink\u003e\\n\\n\u003c/think\u003e\\n\\n' }}\n    {%- else %}\n        {{- '\u003cthink\u003e\\n' }}\n    {%- endif %}\n{%- endif %}","files":[{"name":"model.safetensors.index.json","sha256":"77042094076611b69791a610065f28b7013b8c621795fa86ddccc8bac7d1b9df","size":112216},{"name":"model-00001-of-00018.safetensors","sha256":"ba0ce20aae489ad196733da5064bcdf159a1fe84f53336648196e1ebb7751b1c","size":3966730552},{"name":"model-00002-of-00018.safetensors","sha256":"06a148c01bfbe3faa14a5f184a7ff29a706f7ae1c8b2705d2058e26d17a001fb","size":3043080328},{"name":"model-00003-of-00018.safetensors","sha256":"2e1bf62cbcd406eaa64b60d10353e1f0ef4039d0976e56f05cabe953454f9968","size":2542796952},{"name":"model-00004-of-00018.safetensors","sha256":"511e34063187882659753c4d93f3859f93c019fd438d8813071921c81d9a3f1a","size":3988973152},{"name":"model-00005-of-00018.safetensors","sha256":"635cb53446dc74f219740fc59e18b774f877b803b9722e289ca62575a6efa701","size":2099339864},{"name":"model-00006-of-00018.safetensors","sha256":"0bc5214fac607f0e6cc92eec3789d4b8559410ef9fce66621ba8158e8410dae0","size":3979553696},{"name":"model-00007-of-00018.safetensors","sha256":"80b0c49033e9a0d5762562aa12f4acdb7f54da586f3d0110f28c48d91cf07892","size":2108759344},{"name":"model-00008-of-00018.safetensors","sha256":"7192c5b66185d3592927daabee1cc19e6f6e0ce75988ee20e824b624765fda79","size":3979553696},{"name":"model-00009-of-00018.safetensors","sha256":"af3c48cc37af44f3db6ae0579baf019180d48d9c527caa0a1f03ff85813a56d8","size":2108759344},{"name":"model-00010-of-00018.safetensors","sha256":"163490a76f3bea3a40855b7efc04ce6d27afaf1a34f0bbde495b9491f76457c9","size":3979553696},{"name":"model-00011-of-00018.safetensors","sha256":"5f3ae1b948aeee39da77aec558e8236cd65fe4d7cb7686a76bb007acc563c6d8","size":2108759344},{"name":"model-00012-of-00018.safetensors","sha256":"a3de1c7114677a8f5ac5c4892c90e8238ea5c1e2038c80e757dfc87c3902ca55","size":3979553696},{"name":"model-00013-of-00018.safetensors","sha256":"06ab79a41f74c9c5cb734816feb0c7fc364104b227165ee7391231e1155aa02a","size":2108759344},{"name":"model-00014-of-00018.safetensors","sha256":"4138ed94603065ba884bbcadedb04d7718bb40117e85e6f5c6fc5b9c05b7a85b","size":3979553696},{"name":"model-00015-of-00018.safetensors","sha256":"69224e27b9de4e7dbf6fc936c6eaae08447bda3b80a6c31a871ab451173afd22","size":2108759344},{"name":"model-00016-of-00018.safetensors","sha256":"73cb9a1089fb6155cb648609478d6633be8a5c7d9ca5a05bc8925ce8a553cefe","size":3979564040},{"name":"model-00017-of-00018.safetensors","sha256":"beb51f01056142ac4984bd800507b0dd0fd18de57f8e9ef6ea41d1a3598983a8","size":2108759344},{"name":"model-00018-of-00018.safetensors","sha256":"1d3479509e21494658f9b64d317f5ea8e55c4025d28c702d6c4d0b356ce8ea06","size":3392197344}],"model_size":27781427952,"tensor_type":["BF16"]}},"ModelRevisions":null,"ModelSource":"USER_UPLOAD","ModelTools":"","ModelType":["qwen3_5"],"MuseInfo":null,"NEXA":{"Catalogues":null,"ModelCover":"","ScientificField":"","Source":"","SubScientificField":null},"Name":"Qwen3.8-27B","NewVersion":"","NickName":"yingda","OfficialTags":null,"OpenAiSwingDeployInfo":{"Order":0,"Recommend":null,"lmdeploy":{"eas":{"Script":"","requirements":""},"ens":{"Script":"","requirements":""},"fc":{"Script":"","requirements":""},"image_tag":""},"ollama":{"eas":{"Script":"","requirements":""},"ens":{"Script":"","requirements":""},"fc":{"Script":"","requirements":""},"image_tag":""},"pipeline":{"eas":{"Script":"","requirements":""},"ens":{"Script":"","requirements":""},"fc":{"Script":"","requirements":""},"image_tag":""},"vllm":{"eas":{"Script":"","requirements":""},"ens":{"Script":"","requirements":""},"fc":{"Script":"","requirements":""},"image_tag":""}},"Organization":{"ApplyFailureReason":"","ApplyReason":"","Avatar":"https://resources.modelscope.cn/avatar/4c40e6ce-1348-43b5-a3bc-6bafec6ac805.jpg","CreateCompetition":false,"CreatedBy":"Cherrytest","Description":"[\"root\",{},[\"p\",{},[\"span\",{\"data-type\":\"text\"},[\"span\",{\"data-type\":\"leaf\"},\"欢迎来到 Qwen 👋\"]]],[\"p\",{},[\"span\",{\"data-type\":\"text\"},[\"span\",{\"data-type\":\"leaf\"},\"这是 Qwen 的组织，阿里云构建的大型语言模型家族。在这个组织中，我们不断发布大型语言模型 (LLM)、大型多模态模型 (LMM) 和其他 AGI 相关项目。快来查看并享受吧！\"]]]]","DisplayUrl":"","Email":"","FromSite":"","FullName":"千问","GithubAddress":"https://github.com/QwenLM","GmtCreated":"2023-08-02T03:48:45Z","GmtModified":"2026-06-23T07:20:39Z","Id":209,"InitAdminMembers":"","IsApply":false,"IsCertification":"","Mobile":"","Name":"Qwen","Path":"","Roles":null,"StarCnt":0,"Status":0,"SubscribeVo":null,"Type":2},"PaiModelGalleryUrl":null,"PaiSdkParameter":null,"Path":"Qwen","PreReleaseData":{"AlreadyStar":false,"Description":"新一代Qwen模型：突破极致智能密度","Enabled":false,"Introduce":"# Qwen3.8-27B\n\nQwen3.8 系列在代码编写、专业工作、科学研究及长流程 Agent 任务方面提供顶尖性能。Qwen3.8-27B 将这些技术提升落地于轻量化、易部署的稠密模型：它是一款原生多模态视觉-语言模型，具备图像与视频理解能力，并支持灵活的思考控制（Thinking Control），专为以更高可靠性完整执行复杂的跨步骤任务而设计。\n\n\u003cbr /\u003e\n\n## 亮点\n\n- *Agentic 智能体执行力*: 更强大的自主规划与环境反馈应对能力，实现更高可靠性的端到端任务交付。\n- *可控思考机制*: 提供可开关的思考模式，并支持自定义推理深度（Reasoning Depth）。\n- *多模态视觉理解*: 原生集成图像与视频理解，全面覆盖 STEM 图表、复杂文档及长视频分析。\n\n\u003cbr /\u003e\n\n## 获取最新动态\n\n点击「点赞并获取通知」，模型一经发布，我们将第一时间向您发送通知。届时您将能在本页面获取 Qwen3.8-27B 开源模型权重。\n\n\u003cbr /\u003e\n\n","PlannedArtifacts":["Qwen/Qwen3.8-27B","Qwen/Qwen3.8-27B-FP8"],"ReleaseDate":"2026-08-14T15:00:00.000Z","Stars":865,"SubscriberSample":[{"Avatar":"https://img.alicdn.com/imgextra/i1/O1CN01yhHrHg1Pdl3UKPhGc_!!6000000001864-2-tps-88-88.png","Name":"modelhere"},{"Avatar":"https://img.alicdn.com/imgextra/i1/O1CN01yhHrHg1Pdl3UKPhGc_!!6000000001864-2-tps-88-88.png","Name":"boxscwei"},{"Avatar":"https://img.alicdn.com/imgextra/i1/O1CN01yhHrHg1Pdl3UKPhGc_!!6000000001864-2-tps-88-88.png","Name":"fsss618"},{"Avatar":"https://resources.modelscope.cn/avatar/560da8a2-185c-4bda-aa90-b57144e305ae.png","Name":"TIFcFFF"}],"UiConfig":"{\"modelNameImageUrl\":\"https://resources.modelscope.cn/models/pre-release/7e5f9fcb-10ea-4d49-b1d1-0857de449050.png\",\"modelLogoImageUrl\":\"https://resources.modelscope.cn/models/pre-release/b47786f9-58ec-4b02-aed3-3f93eead8fbf.png\"}"},"ProtectedMode":2,"ReadMeContent":"\n# Qwen3.8-27B\n\n\u003e [!Note]\n\u003e This repository contains model weights and configuration files for the post-trained model in the Hugging Face Transformers format. \n\u003e\n\u003e These artifacts are compatible with Hugging Face Transformers, vLLM, SGLang, TokenSpeed, etc.\n\n\u003e [!Tip]\n\u003e For users seeking managed, scalable inference without infrastructure maintenance, the official Qwen API service is provided by [Qwen Cloud](https://www.qwencloud.com).\n\u003e In particular, **Qwen3.8-27B** will be available as a hosted version with more production features, e.g., 1M context length by default, official built-in tools. For more information, please refer to the [Qwen3.8-27B Overview](https://www.qwencloud.com/models/qwen3.8-27b). The service is coming soon. Stay tuned for updates.\n\nFollowing the widespread community adoption of the Qwen3.5 and Qwen3.6 series, we are pleased to introduce Qwen3.8, the most capable generation in the Qwen open-model family to date.\n\nBuilt on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Qwen3.8-27B brings these advances to a compact, deployment-friendly dense model: a native vision-language model that understands images and videos, with flexible thinking control, designed to carry complex, multi-step tasks through to completion with greater reliability.\n\n## Qwen3.8 Highlights\n\nQwen3.8-27B features the following enhancements:\n- **Core Capabilities**: Comprehensive improvements across coding, professional work, research, and long-horizon agentic tasks.\n- **Agent Execution**: Stronger autonomous planning and better handling of environment feedback, leading to more reliable end-to-end task completion.\n- **Downstream Compatibility**: Broader support for popular harnesses and development tools, making it easier to integrate into your existing stack.\n- **Flexible Thinking Control**: Thinking mode is on by default and can be disabled per request; reasoning depth can be tuned with `reasoning_effort`, and reasoning context from historical messages is retained via `preserve_thinking`.\n- **Vision-Language Understanding**: Native support for image and video understanding, from STEM diagrams and documents to hour-scale videos.\n\n\n## Model Overview\n\n- Type: Causal Language Model with Vision Encoder\n- Training Stage: Pre-training \u0026 Post-training\n- Language Model\n    - Number of Parameters: 27B\n    - Hidden Dimension: 5120\n    - Token Embedding: 248,320 (Padded)\n    - Number of Layers: 64\n    - Hidden Layout: 16 × (3 × (Gated DeltaNet → FFN) → 1 × (Gated Attention → FFN))\n    - Gated DeltaNet:\n        - Number of Linear Attention Heads: 48 for V and 16 for QK\n        - Head Dimension: 128\n    - Gated Attention:\n        - Number of Attention Heads: 24 for Q and 4 for KV\n        - Head Dimension: 256\n        - Rotary Position Embedding Dimension: 64\n    - Feed Forward Network:\n        - Intermediate Dimension: 17,408\n    - LM Output: 248,320 (Padded)\n    - MTP (Multi-Token Prediction): trained with multiple steps\n- Context Length: 262,144 natively and extensible up to 1,000,000 tokens.\n\n\n## Benchmark Results\n\n### Text Performance\n\u003cstyle\u003e\n.vl-table th{font-size:15px!important;line-height:1.2}\n.vl-table td:not(.benchmark-cell):not([colspan]){font-size:15px;line-height:1.2;vertical-align:middle}\n.vl-table .benchmark-cell{padding:12px 10px 12px 18px!important;vertical-align:middle}\n.vl-table .benchmark-capability{font-size:15px;font-weight:600;line-height:1.22;color:#171717}\n.vl-table .benchmark-name{margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B}\n.vl-table .metric-stack{display:flex;flex-direction:column;gap:7px;padding:3px 0}\n.vl-table .metric-label{font-size:10px;font-weight:400;line-height:1.1;color:#777}\n.vl-table .metric-value{margin-top:2px;font-size:15px;line-height:1.15;color:#171717}\n\u003c/style\u003e\n\u003cdiv style=\"font-family:-apple-system,BlinkMacSystemFont,'Segoe UI',Roboto,sans-serif;max-width:1200px;margin:0 auto;padding:16px 0\"\u003e\n\u003ctable class=\"vl-table\" style=\"width:100%;table-layout:fixed;border-collapse:collapse;font-size:13px\"\u003e\n\u003cthead\u003e\u003ctr\u003e\n\u003cth style=\"padding:10px 7px;text-align:left;font-weight:600;border-bottom:2px solid #0A2EFE;color:#0A2EFE\"\u003e\u003c/th\u003e\u003cth style=\"padding:10px 7px;text-align:center;font-weight:500;border-bottom:2px solid #0A2EFE;color:#0A2EFE;font-size: 14px;width:14.00%;background:rgba(10, 46, 254, 0.08);\"\u003eQwen3.8-27B\u003c/th\u003e\u003cth style=\"padding:10px 7px;text-align:center;font-weight:500;border-bottom:2px solid #0A2EFE;color:#0A2EFE;font-size: 14px;width:14.00%;\"\u003eQwen3.6-27B\u003c/th\u003e\u003cth style=\"padding:10px 7px;text-align:center;font-weight:500;border-bottom:2px solid #0A2EFE;color:#0A2EFE;font-size: 14px;width:14.00%;\"\u003eQwen3.7-Plus\u003c/th\u003e\u003cth style=\"padding:10px 7px;text-align:center;font-weight:500;border-bottom:2px solid #0A2EFE;color:#0A2EFE;font-size: 14px;width:14.00%;\"\u003eMuse Glimmer-30B\u003c/th\u003e\u003cth style=\"padding:10px 7px;text-align:center;font-weight:500;border-bottom:2px solid #0A2EFE;color:#0A2EFE;font-size: 14px;width:14.00%;\"\u003eOpus4.6 Max\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\u003ctd colspan=\"6\" style=\"padding:8px 12px;font-weight:600;color:#0A2EFE;border-bottom:1px solid rgba(10, 46, 254, 0.2);background:#D6DAFC\"\u003eCoding\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eAgentic terminal coding\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eTerminal Bench 2.1 (Terminus)\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e73.0\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e63.4\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e64.0\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e51.7\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e78.2\u003c/strong\u003e\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eAgentic coding\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eSWE-bench Pro\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e61.7\u003c/strong\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e53.5\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e57.6\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e51.2\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e53.4\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eRepo-level code generation\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eNL2Repo-Bench\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e42.3\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e36.2\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e41.1\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e47.6\u003c/strong\u003e\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eAgentic coding\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eDeepSWE 1.1\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e42.2\u003c/strong\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e13.3\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e14.2\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eSoftware engineering\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eQwenSWEBench\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e79.0\u003c/strong\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e49.3\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e59.2\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e63.8\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\u003ctd colspan=\"6\" style=\"padding:8px 12px;font-weight:600;color:#0A2EFE;border-bottom:1px solid rgba(10, 46, 254, 0.2);background:#D6DAFC\"\u003eAgent\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eLong-horizon office work\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eCoWorkBench\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e70.7\u003c/strong\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e61.0\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e65.1\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e68.2\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eProfessional job tasks\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eJobBench\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e33.4\u003c/strong\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e21.8\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e27.6\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eFrontier agentic tasks\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eAgents' Last Exam\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003ePass@1\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e20.4\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eScore\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e42.9\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003ePass@1\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e10.6\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eScore\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e27.3\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003ePass@1\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e13.2\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eScore\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e33.6\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\u003ctd colspan=\"6\" style=\"padding:8px 12px;font-weight:600;color:#0A2EFE;border-bottom:1px solid rgba(10, 46, 254, 0.2);background:#D6DAFC\"\u003eGeneral\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eInstruction following\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eIFBench\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e79.5\u003c/strong\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e69.1\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e79.1\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e77.0\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e62.5\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eScientific reasoning\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eGPQA Diamond\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e89.2\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e87.8\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e90.3\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e83.5\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e91.3\u003c/strong\u003e\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eMultidisciplinary reasoning\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eHLE\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e30.8\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e24.0\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e34.7\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e22.0\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e40.0\u003c/strong\u003e\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eCompetitive coding\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eLiveCodeBench v6\u003c/div\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e90.3\u003c/strong\u003e\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e83.9\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e89.6\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\n\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e88.8\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003cdiv style=\"margin-top:12px;font-size:11px;line-height:1.5;color:rgba(0,0,0,0.72)\"\u003e\n\u003col style=\"margin:0;padding-left:20px\"\u003e\n\u003cli\u003eSWE-bench Pro: Except for Opus4.6 Max, which uses the officially reported score, all models are evaluated with the Claude Code harness at temp=1.0, top_p=0.95, and a 256K context window. Problematic tasks were corrected, and all baseline models were re-evaluated on the refined benchmark.\u003c/li\u003e\n\u003cli\u003eNL2Repo-Bench: Evaluated with the Claude Code harness. To prevent reward hacking, we disable Bash commands that attempt to access the specific repository, such as pip download, pip install, and git clone.\u003c/li\u003e\n\u003cli\u003eDeepSWE 1.1: Evaluated with the Claude Code harness at temp=1.0, top_p=0.95, and a 256K context window.\u003c/li\u003e\n\u003cli\u003eQwenSWEBench: In-house coding benchmark for evaluating models' software engineering capabilities. Evaluated with the Claude Code harness. Reporting avg@3 with an 8-hour timeout, max_tokens=32,768, temperature=1.0, and a 256K context window.\u003c/li\u003e\n\u003cli\u003eCoWorkBench: In-house cowork benchmark for evaluating long-horizon tasks across computer science, finance, law, medical, and other productivity domains.\u003c/li\u003e\n\u003cli\u003eHLE: Judged by GPT-4o.\u003c/li\u003e\n\u003cli\u003eThe best result in each row is shown in bold.\u003c/li\u003e\n\u003cli\u003eEmpty cells (--) indicate that results are not yet available or not applicable.\u003c/li\u003e\n\u003c/ol\u003e\n\u003c/div\u003e\n\u003c/div\u003e\n\n### VL Performance\n\u003cdiv style=\"font-family:-apple-system,BlinkMacSystemFont,'Segoe UI',Roboto,sans-serif;max-width:1200px;margin:0 auto;padding:16px 0\"\u003e\n\u003ctable class=\"vl-table\" style=\"width:100%;table-layout:fixed;border-collapse:collapse;font-size:13px\"\u003e\n\u003cthead\u003e\u003ctr\u003e\u003cth style=\"padding:10px 7px;text-align:left;font-weight:600;border-bottom:2px solid #0A2EFE;color:#0A2EFE\"\u003e\u003c/th\u003e\u003cth style=\"padding:10px 7px;text-align:center;font-weight:500;border-bottom:2px solid #0A2EFE;color:#0A2EFE;font-size: 14px;width:14.00%;background:rgba(10, 46, 254, 0.08);\"\u003eQwen3.8-27B\u003c/th\u003e\u003cth style=\"padding:10px 7px;text-align:center;font-weight:500;border-bottom:2px solid #0A2EFE;color:#0A2EFE;font-size: 14px;width:14.00%;\"\u003eQwen3.6-27B\u003c/th\u003e\u003cth style=\"padding:10px 7px;text-align:center;font-weight:500;border-bottom:2px solid #0A2EFE;color:#0A2EFE;font-size: 14px;width:14.00%;\"\u003eQwen3.7-Plus\u003c/th\u003e\u003cth style=\"padding:10px 7px;text-align:center;font-weight:500;border-bottom:2px solid #0A2EFE;color:#0A2EFE;font-size: 14px;width:14.00%;\"\u003eMuse Glimmer-30B\u003c/th\u003e\u003cth style=\"padding:10px 7px;text-align:center;font-weight:500;border-bottom:2px solid #0A2EFE;color:#0A2EFE;font-size: 14px;width:14.00%;\"\u003eOpus4.6 Max\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\u003ctd colspan=\"6\" style=\"padding:8px 12px;font-weight:600;color:#0A2EFE;border-bottom:1px solid rgba(10, 46, 254, 0.2);background:#D6DAFC\"\u003eAgentic Multimodal Intelligence\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eComputer use\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eOSWorld-Verified\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e84.3\u003c/strong\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e63.9\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e73.3\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e65.9\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e72.7\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eBrowser use\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eWebArena-Verified\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e64.8\u003c/strong\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e48.8\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e55.3\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eMobile use\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eAndroidWorld\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e81.9\u003c/strong\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e70.3\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e81.0\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e62.0\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eApplication recreation\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eRecreationBench\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e47.1\u003c/strong\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e29.8\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e30.2\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eMultimodal tool use\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eClawEval-MM\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003ePass@3\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e57.4\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eAverage\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e56.9\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003ePass@3\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e42.6\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eAverage\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e50.4\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003ePass@3\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e57.4\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eAverage\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e60.1\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003ePass@3\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e52.5\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eAverage\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e54.7\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eMultimodal software engineering\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eSWE-MM\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e38.6\u003c/strong\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e25.7\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e30.0\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e27.1\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eVisual web development\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eVision2Web\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e62.9\u003c/strong\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e45.0\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e42.1\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd colspan=\"6\" style=\"padding:8px 12px;font-weight:600;color:#0A2EFE;border-bottom:1px solid rgba(10, 46, 254, 0.2);background:#D6DAFC\"\u003eGeneral Multimodal Intelligence\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eVisual math problem solving\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eMathVision\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e90.0\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWith CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e94.6\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e85.1\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e90.3\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e65.5\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eGeneral visual reasoning\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eBabyVision\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e65.7\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWith CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e85.6\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e28.9\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e64.7\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWith CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e70.4\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e12.6\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eScientific chart analysis\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eCharXiv (RQ)\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e83.7\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWith CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e90.2\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e78.4\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e\u003cstrong\u003e85.8\u003c/strong\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv style=\"margin-top:7px\"\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWith CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e85.9\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e78.8\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cdiv class=\"metric-stack\" style=\"padding:3px 0\"\u003e\u003cdiv\u003e\u003cdiv class=\"metric-label\" style=\"font-size:10px;font-weight:400;line-height:1.1;color:#777\"\u003eWithout CI\u003c/div\u003e\u003cdiv class=\"metric-value\" style=\"margin-top:2px;font-size:15px;line-height:1.15;color:#171717\"\u003e66.0\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eDocument intelligence\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eOmniDocBench 1.5\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e91.1\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e89.4\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e91.4\u003c/strong\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e75.8\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e86.6\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eReal-world perception\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eRealWorldQA\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e85.9\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e84.1\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e86.9\u003c/strong\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e73.9\u003c/td\u003e\u003c/tr\u003e\n\u003ctr\u003e\u003ctd class=\"benchmark-cell\" style=\"padding:7px 7px;padding-left:20px;border-bottom:1px solid rgba(128, 128, 128, 0.15);\"\u003e\u003cdiv class=\"benchmark-capability\" style=\"font-size:15px;font-weight:600;line-height:1.22;color:#171717\"\u003eEmbodied intelligence\u003c/div\u003e\u003cdiv class=\"benchmark-name\" style=\"margin-top:4px;font-size:11px;font-weight:400;line-height:1.2;color:#6B6B6B\"\u003eERQA\u003c/div\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);background:rgba(10, 46, 254, 0.08);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e65.5\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e62.5\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e\u003cstrong\u003e69.8\u003c/strong\u003e\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e--\u003c/td\u003e\u003ctd style=\"padding:7px 7px;text-align:center;border-bottom:1px solid rgba(128, 128, 128, 0.15);vertical-align:middle;font-size:15px;line-height:1.2;\"\u003e40.8\u003c/td\u003e\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003cdiv style=\"margin-top:12px;font-size:11px;line-height:1.5;color:rgba(0,0,0,0.72)\"\u003e\n\u003col style=\"margin:0;padding-left:20px\"\u003e\n\u003cli\u003eMathVision, BabyVision, and CharXiv (RQ): Where both settings are available, cells report “Without CI” and “With CI” separately; otherwise, only the available setting is shown. A small number of incorrect ground-truth annotations in MathVision and CharXiv (RQ) were corrected following manual verification, and all reported scores on those benchmarks were computed using the corrected annotations.\u003c/li\u003e\n\u003cli\u003eMathVision: Qwen3.8-27B is evaluated using the fixed prompt: “Please reason step by step, and put your final answer within \u003ccode\u003e\\boxed{}\u003c/code\u003e.” For the remaining models, we report the higher score from two prompt variants—one with and one without the \u003ccode\u003e\\boxed{}\u003c/code\u003e formatting requirement.\u003c/li\u003e\n\u003cli\u003eWebArena-Verified: Scores are computed with the official WebArena-Verified grader under the OSWorld scaffold.\u003c/li\u003e\n\u003cli\u003eRecreationBench: An in-house, long-horizon application-recreation benchmark designed to evaluate hybrid-agent capabilities across five platforms: desktop (Ubuntu, macOS, and Windows), mobile (Android), and the web.\u003c/li\u003e\n\u003cli\u003eClawEval-MM: Scores are reported as “Pass@3 / average score.” Pass@3 is the percentage of tasks passed in at least one of three trials; the average score is the mean benchmark score across the three trials.\u003c/li\u003e\n\u003cli\u003eVision2Web: Scores are averaged across the frontend, webpage, and website categories. Evaluations use the Claude Code harness and are judged by \u003ccode\u003egpt-5.4-2026-03-05\u003c/code\u003e.\u003c/li\u003e\n\u003cli\u003eSWE-MM: Scores are evaluated on the Claude Code harness using the public dev split of SWE-bench Multimodal, with the modifications described in Appendix 8.3 of the Claude Opus 4.7 system card.\u003c/li\u003e\n\u003cli\u003eEmpty cells (--) indicate that results are not yet available or not applicable.\u003c/li\u003e\n\u003c/ol\u003e\u003c/div\u003e\n\u003c/div\u003e\n\n\n## Quickstart\n\nFor streamlined integration, we recommend using Qwen3.8 via APIs.\n\n### Serving Qwen3.8\n\n\u003e [!Important]\n\u003e Inference efficiency and throughput vary significantly across frameworks. \n\u003e We recommend using the latest framework versions to ensure optimal performance and compatibility.\n\u003e For production workloads or high-throughput scenarios, dedicated serving engines such as SGLang, vLLM, or TokenSpeed are recommended.\n\nQwen3.8 can be deployed with popular inference frameworks, e.g.:\n\n- [SGLang](https://www.sglang.io/): [Qwen3.8 Cookbook](https://docs.sglang.io/cookbook/autoregressive/Qwen/Qwen3.8-27B)\n- [vLLM](https://vllm.ai/): [Qwen3.8 Recipe](https://recipes.vllm.ai/Qwen/Qwen3.8-27B)\n- [TokenSpeed](https://lightseek.org/tokenspeed/): [Qwen3.8 Recipe](https://lightseek.org/tokenspeed/recipes/models#qwen3-8)\n\n\n### API Usage\n\n\u003e [!Important]\n\u003e Qwen3.8 models operate in thinking mode by default, generating thinking content signified by `\u003cthink\u003e\\n...\u003c/think\u003e\\n\\n` before producing the final response.\n\u003e To disable thinking content and obtain a direct response, refer to the examples [here](#instruct-or-non-thinking-mode).\n\n\n\u003e [!Tip]\n\u003e We recommend using the following sets of sampling parameters for generation:\n\u003e - Thinking Mode: `temperature=1.0`, `top_p=0.95`, `top_k=20`, `min_p=0.0`, `presence_penalty=0.0`, `repetition_penalty=1.0`\n\u003e - Instruct (or non-thinking) mode: `temperature=0.7`, `top_p=0.80`, `top_k=20`, `min_p=0.0`, `presence_penalty=1.5`, `repetition_penalty=1.0`\n\u003e\n\u003e Please note that the support for sampling parameters varies according to inference frameworks.\n\n\nQwen3.8 comes with official support for `reasoning_effort`, which can be used to adjust reasoning depth and control cost:  \n  - `xhigh` (default): for complex tasks demanding thorough analysis\n  - `medium`: balancing accuracy and speed\n  - `low`: efficient reasoning optimizing for speed and cost\n\n\nIn addition, `preserve_thinking` is enabled by default for all workloads for the best out-of-the-box experience. To disable preserved thinking, refer to the examples [here](#disable-preserved-thinking).\n\n\u003e [!Tip]\n\u003e In multi-turn agentic tasks, lower reasoning effort does not always reduce overall task completion time. Although it may produce faster per-turn responses, it can also lead to insufficient analysis, more failures, and repeated retries, which may increase total latency and token consumption.\n\n\n#### Chat Completions API\n\nThe Chat Completions API can be used with most inference frameworks, as well as [Qwen Cloud](https://www.qwencloud.com/).\nBefore starting, make sure the OpenAI Python SDK is installed and the API key and the API base URL are configured, e.g.:\n```shell\npip install -U openai\n\n# Set the following accordingly\nexport OPENAI_BASE_URL='your-base-url'\nexport OPENAI_API_KEY='your-api-key'\n```\n\n##### Text-Only Input\n\n```python\nfrom openai import OpenAI\n# Configured by environment variables\nclient = OpenAI()\n\nmessages = [{\"role\": \"user\", \"content\": \"Write a Python function to merge two sorted linked lists.\"}]\n\ncompletion = client.chat.completions.create(\n    model=\"Qwen/Qwen3.8-27B\",\n    messages=messages,\n    extra_body={\n        \"chat_template_kwargs\": {\n            \"enable_thinking\": True,  # on by default\n            \"preserve_thinking\": True, # on by default\n        },\n    },\n    reasoning_effort=\"xhigh\",  # xhigh by default; supported levels are xhigh, medium, and low\n    stream=True,\n    stream_options={\"include_usage\": True},\n)\n\nreasoning_content = \"\"\nanswer_content = \"\"\nis_answering = False\nprint(\"\\n\" + \"=\" * 20 + \"Reasoning\" + \"=\" * 20 + \"\\n\")\n\nfor chunk in completion:\n    if not chunk.choices:\n        print(\"\\nUsage:\")\n        print(chunk.usage)\n        continue\n\n    delta = chunk.choices[0].delta\n\n    if hasattr(delta, \"reasoning_content\") and delta.reasoning_content is not None:\n        if not is_answering:\n            print(delta.reasoning_content, end=\"\", flush=True)\n        reasoning_content += delta.reasoning_content\n    elif hasattr(delta, \"reasoning\") and delta.reasoning is not None:\n        if not is_answering:\n            print(delta.reasoning, end=\"\", flush=True)\n        reasoning_content += delta.reasoning\n\n    if hasattr(delta, \"content\") and delta.content:\n        if not is_answering:\n            print(\"\\n\" + \"=\" * 20 + \"Answer\" + \"=\" * 20 + \"\\n\")\n            is_answering = True\n        print(delta.content, end=\"\", flush=True)\n        answer_content += delta.content\n\nmessages.append({\n    \"role\": \"assistant\",\n    \"content\": answer_content,\n    \"reasoning_content\": reasoning_content,\n    \"reasoning\": reasoning_content,\n})\n```\n\n\n##### Image Input\n\n```python\nfrom openai import OpenAI\n# Configured by environment variables\nclient = OpenAI()\n\nmessages = [\n    {\n        \"role\": \"user\",\n        \"content\": [\n            {\n                \"type\": \"image_url\",\n                \"image_url\": {\n                    \"url\": \"https://qianwen-res.oss-accelerate.aliyuncs.com/Qwen3.5/demo/CI_Demo/mathv-1327.jpg\"\n                }\n            },\n            {\n                \"type\": \"text\",\n                \"text\": \"The centres of the four illustrated circles are in the corners of the square. The two big circles touch each other and also the two little circles. With which factor do you have to multiply the radii of the little circles to obtain the radius of the big circles?\\nChoices:\\n(A) $\\\\frac{2}{9}$\\n(B) $\\\\sqrt{5}$\\n(C) $0.8 \\\\cdot \\\\pi$\\n(D) 2.5\\n(E) $1+\\\\sqrt{2}$\"\n            }\n        ]\n    }\n]\n\nchat_response = client.chat.completions.create(\n    model=\"Qwen/Qwen3.8-27B\",\n    messages=messages,\n)\nprint(\"Chat response:\", chat_response)\n```\n\n##### Video Input\n\n```python\nfrom openai import OpenAI\n# Configured by environment variables\nclient = OpenAI()\n\nmessages = [\n    {\n        \"role\": \"user\",\n        \"content\": [\n            {\n                \"type\": \"video_url\",\n                \"video_url\": {\n                    \"url\": \"https://qianwen-res.oss-accelerate.aliyuncs.com/Qwen3.5/demo/video/N1cdUjctpG8.mp4\"\n                }\n            },\n            {\n                \"type\": \"text\",\n                \"text\": \"How many porcelain jars were discovered in the niches located in the primary chamber of the tomb?\"\n            }\n        ]\n    }\n]\n\nchat_response = client.chat.completions.create(\n    model=\"Qwen/Qwen3.8-27B\",\n    messages=messages,\n)\n\n# When vLLM is launched with `--media-io-kwargs '{\"video\": {\"num_frames\": -1}}'`,\n# video frame sampling can be configured via `extra_body` (e.g., by setting `fps`).\n# This feature is currently supported only in vLLM.\n#\n# By default, `fps=2` and `do_sample_frames=True`.\n# With `do_sample_frames=True`, you can customize the `fps` value to set your desired video sampling rate.\n# chat_response = client.chat.completions.create(\n#     model=\"Qwen/Qwen3.8-27B\",\n#     messages=messages,\n#     extra_body={\n#         \"mm_processor_kwargs\": {\"fps\": 2, \"do_sample_frames\": True},\n#     }, \n# )\n\nprint(\"Chat response:\", chat_response)\n```\n\n\n##### Instruct (or Non-Thinking) Mode\n\nQwen3.8-27B will think by default before responding.\nYou can obtain a direct response from the model without thinking by configuring the API parameters. \nFor example,\n```python\nfrom openai import OpenAI\n# Configured by environment variables\nclient = OpenAI()\n\nmessages = [\n    {\n        \"role\": \"user\",\n        \"content\": [\n            {\n                \"type\": \"image_url\",\n                \"image_url\": {\n                    \"url\": \"https://qianwen-res.oss-accelerate.aliyuncs.com/Qwen3.5/demo/RealWorld/RealWorld-04.png\"\n                }\n            },\n            {\n                \"type\": \"text\",\n                \"text\": \"Where is this?\"\n            }\n        ]\n    }\n]\n\nchat_response = client.chat.completions.create(\n    model=\"Qwen/Qwen3.8-27B\",\n    messages=messages,\n    temperature=0.7,\n    top_p=0.8,\n    presence_penalty=1.5,\n    extra_body={\n        \"top_k\": 20,\n        \"chat_template_kwargs\": {\"enable_thinking\": False},\n    }, \n)\nprint(\"Chat response:\", chat_response)\n```\n\n\u003e [!Note]\n\u003e If you are using APIs from Qwen Cloud, in addition to changing `model`, please use `\"enable_thinking\": False` instead of `\"chat_template_kwargs\": {\"enable_thinking\": False}`.\n\n\n##### Disable Preserved Thinking\n\n\nBy default, Qwen3.8 retains thinking blocks from all historical messages, maintaining a complete reasoning trace across the conversation. This behavior, known as preserved thinking, ensures full context continuity and is especially beneficial for agent scenarios where decision consistency and reduced redundant reasoning are critical. It also improves KV cache utilization, optimizing inference efficiency in both thinking and non-thinking modes.\n\nIf you prefer to retain only the thinking blocks from the latest user message, you can disable this behavior by setting `preserve_thinking` to `False`:\n\n```python\nfrom openai import OpenAI\n\n# Configured by environment variables\nclient = OpenAI()\nmessages = [...]\nchat_response = client.chat.completions.create(\n    model=\"Qwen/Qwen3.8-27B\",\n    messages=messages,\n    extra_body={\n        \"chat_template_kwargs\": {\"preserve_thinking\": False},\n    },\n)\nprint(\"Chat response:\", chat_response)\n```\n\n\u003e [!Note]\n\u003e If you are using APIs from Qwen Cloud, in addition to changing `model`, please use `\"preserve_thinking\": False` directly instead of wrapping it in `chat_template_kwargs`.\n\n\n## Best Practices\n\nTo achieve optimal performance, we recommend the following settings:\n\n1. **Sampling Parameters**: We suggest using the following sets of sampling parameters:  \n    \n    - Thinking Mode: `temperature=1.0`, `top_p=0.95`, `top_k=20`, `min_p=0.0`, `presence_penalty=0.0`, `repetition_penalty=1.0`\n    - Instruct (or non-thinking) mode: `temperature=0.7`, `top_p=0.80`, `top_k=20`, `min_p=0.0`, `presence_penalty=1.5`, `repetition_penalty=1.0`\n    \n    For supported frameworks, you can adjust the `presence_penalty` parameter between 0 and 2 to reduce endless repetition. However, using a higher value may occasionally result in language mixing and a slight decrease in model performance.\n\n2. **Adequate Output Length**: To optimize performance on agentic tasks, we recommend allocating sufficient output length to allow the model to generate detailed and comprehensive responses. For frameworks that support separate token limits for internal reasoning and final outputs, we suggest the following configuration within the 1M context length:\n    \n    - Reasoning Content: Set the maximum output length to 262,144 tokens.\n    - Final Response: Set the maximum output length to 131,072 tokens.\n\n    These settings provide the necessary capacity for complex reasoning while ensuring ample space for high-quality final deliverables.\n\n3. **Processing Ultra-Long Texts**: Qwen3.8-27B natively supports context lengths of up to 262,144 tokens. For long-horizon tasks where the total length (including both input and output) exceeds this limit, we recommend using RoPE scaling techniques to handle long texts effectively, e.g., YaRN.\n\n    YaRN is currently supported by several inference frameworks, e.g., vLLM, SGLang, and TokenSpeed. \n    In general, there are two approaches to enabling YaRN for supported frameworks:\n\n    - Modifying the model configuration file:\n        \n        In the `config.json` file, change the `rope_parameters` fields in `text_config` to:\n        ```json\n        {\n            \"mrope_interleaved\": true,\n            \"mrope_section\": [\n                11,\n                11,\n                10\n            ],\n            \"rope_type\": \"yarn\",\n            \"rope_theta\": 10000000,\n            \"partial_rotary_factor\": 0.25,\n            \"factor\": 4.0,\n            \"original_max_position_embeddings\": 262144,\n        }\n        ```\n\n    - Passing command line arguments:\n\n        For vLLM, you can use\n        ```shell\n        VLLM_ALLOW_LONG_MAX_MODEL_LEN=1 vllm serve ... --hf-overrides '{\"text_config\": {\"rope_parameters\": {\"mrope_interleaved\": true, \"mrope_section\": [11, 11, 10], \"rope_type\": \"yarn\", \"rope_theta\": 10000000, \"partial_rotary_factor\": 0.25, \"factor\": 4.0, \"original_max_position_embeddings\": 262144}}}' --max-model-len 1000000  \n        ```\n\n        For SGLang, you can use\n        ```shell\n        SGLANG_ALLOW_OVERWRITE_LONGER_CONTEXT_LEN=1 python -m sglang.launch_server ... --json-model-override-args '{\"text_config\": {\"rope_parameters\": {\"mrope_interleaved\": true, \"mrope_section\": [11, 11, 10], \"rope_type\": \"yarn\", \"rope_theta\": 10000000, \"partial_rotary_factor\": 0.25, \"factor\": 4.0, \"original_max_position_embeddings\": 262144}}}' --context-length 1000000\n        ```\n\n        For TokenSpeed, you can use\n        ```shell\n        TOKENSPEED_ALLOW_OVERWRITE_LONGER_CONTEXT_LEN=1 tokenspeed serve ... --hf-overrides '{\"text_config\": {\"rope_parameters\": {\"mrope_interleaved\": true, \"mrope_section\": [11, 11, 10], \"rope_type\": \"yarn\", \"rope_theta\": 10000000, \"partial_rotary_factor\": 0.25, \"factor\": 4.0, \"original_max_position_embeddings\": 262144}}}' --max-model-len 1000000  \n        ```\n    \n    \u003e [!NOTE]\n    \u003e All the notable open-source frameworks implement static YaRN, which means the scaling factor remains constant regardless of input length, **potentially impacting performance on shorter texts.**\n    \u003e We advise modifying the `rope_parameters` configuration only when processing long contexts is required. \n    \u003e It is also recommended to modify the `factor` as needed. For example, if the typical context length for your application is 524,288 tokens, it would be better to set `factor` as 2.0. \n\n\n4. **Long Video Understanding**: To optimize inference efficiency for plain text and images, the `size` parameter in the released `video_preprocessor_config.json` is conservatively configured. It is recommended to set the `longest_edge` parameter in the video_preprocessor_config file to 469,762,048 (corresponding to 224k video tokens) to enable higher frame-rate sampling for hour-scale videos and thereby achieve superior performance. For example,\n    ```json\n    {\"longest_edge\": 469762048, \"shortest_edge\": 4096}\n    ```\n\n    Alternatively, override the default values via engine startup parameters. For implementation details, refer to: [vLLM](https://github.com/vllm-project/vllm/pull/34330) / [SGLang](https://github.com/sgl-project/sglang/pull/18467).\n\n\n## Citation\n\nIf you find our work helpful, feel free to give us a cite.\n\n\n```bibtex\n@misc{qwen38,\n    title = {{Qwen3.8-Max}: A New Bar for Coding and Cowork},\n    url = {https://qwen.ai/blog?id=qwen3.8},\n    author = {{Qwen Team}},\n    month = {August},\n    year = {2026}\n}\n```\n\n","ReadMeTips":null,"RelatedArxivId":[],"RelatedPaper":[],"Revision":"master","Stars":865,"StorageSize":55586114768,"Studios":[],"SubVisionFoundation":"","SupportApiInference":false,"SupportDashDeployment":0,"SupportDashInference":0,"SupportDashTraining":0,"SupportDeployment":0,"SupportExperience":0,"SupportFinetuning":0,"SupportFlexTrain":0,"SupportInference":"","SupportPaiModelGallery":null,"SupportPaiSdk":0,"SupportPivotEval":false,"SwingDeployInfo":null,"Tags":[],"Tasks":[{"ChineseName":"视觉多模态理解","Description":"","DomainName":"multi-modal","Id":295,"IsExhibition":true,"IsHot":0,"IsLeaf":true,"IsLoginRequired":true,"IsRetrieval":true,"Level":1,"Name":"image-text-to-text","ParentId":-1,"ParentTask":null,"Sorting":0,"SupportWidgets":false,"TypicalModel":"","WidgetConfig":"{}","WidgetValidator":""}],"Tools":[],"TriggerWords":null,"Visibility":5,"VisionFoundation":"","_":null,"widgets":[]},"Message":"success","RequestId":"e73f4774-d20c-49cd-9b38-93097590f037","Success":true}