Qwen3.5 27B
Stored tensors can include buffers and repeated shared weights. A verified unique parameter count is unavailable. Checkpoint size is not the memory required to run the model.
GeekRadar / The comparison desk
COMPARE.
AI modelsQwen3.5 27B
AI model Released
- Context window
- 262,144 tokens
- Maximum output
- 65,536 tokens
- Open weights
- Yes
Capability & preference
Human preference · Arena
Overall text preferences from head-to-head votes. Not factual accuracy or a task-success rate.
Benchmark results
- AI2D_TEST · version unreported · reported score92.90*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- AIME26 · version unreported · reported score · Full AIME 2026 I and II92.60*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Task set
- Full AIME 2026 I and II*
- C-Eval · version unreported · reported score90.50*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- CC-OCR · version unreported · reported score81*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- CharXiv(RQ) · version unreported · reported score79.50*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Claw-Eval Avg · version unreported · reported score64.30*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Claw-Eval Pass^3 · version unreported · reported score46.20*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- DeepPlanning · version unreported · reported score22.60*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- EmbSpatialBench · version unreported · reported score84.50*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- GPQA · version unreported · reported score85.50*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- HallusionBench · version unreported · reported score70*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- HLE · version unreported · reported score24.30*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- HMMT Feb 25 · version unreported · reported score92*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- HMMT Feb 26 · version unreported · reported score84.30*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- HMMT Nov 25 · version unreported · reported score89.80*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- IMOAnswerBench · version unreported · reported score79.90*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- LiveCodeBench v6 · 6 · reported score80.70*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- LVBench · version unreported · reported score73.60*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Mathvista(mini) · version unreported · reported score87.80*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MCP-Atlas · version unreported · reported score · Public set68.40*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Judge
- gemini-2.5-pro*
- Task set
- Public set*
- MCPMark · version unreported · reported score · GitHub MCP v0.30.336.30*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- GitHub MCP v0.30.3*
- Truncation
- Playwright responses at 32K tokens*
- MLVU · version unreported · reported score85.90*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MMBench EN-DEV-v1.1 · 1.1 · reported score92.60*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MMLU-Pro · version unreported · reported score86.10*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MMLU-Redux · version unreported · reported score93.20*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MMMU · version unreported · reported score82.30*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MMMU-Pro · version unreported · reported score75*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MVBench · version unreported · reported score74.60*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- NL2Repo · version unreported · reported score · Claude Code27.30*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- Claude Code*
- Variant
- Maximum 900 turns*
- Temperature
- 1.0*
- Top p
- 0.95*
- ODInW13 · version unreported · reported score41.10*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- OmniDocBench1.5 · 1.5 · reported score88.90*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- QwenClawBench · version unreported · reported score · Internal real-user distribution52.20*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Task set
- Internal real-user distribution*
- Temperature
- 0.6*
- Context length
- 256K tokens*
- QwenWebBench · version unreported · reported score · Internal1,068* EloPlatform unreported Compare results
Test conditions
- Metric
- BT/Elo rating*
- Checkpoint version
- Not specified*
- Languages
- English and Chinese*
- Categories
- Seven frontend application categories*
- Judge
- Multimodal automatic judge*
- Task set
- Internal*
- RealWorldQA · version unreported · reported score83.70*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- RefCOCO(avg) · version unreported · reported score90.90*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- RefSpatialBench · version unreported · reported score67.70*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- SimpleVQA · version unreported · reported score56*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- SkillsBench Avg5 · version unreported · reported score · OpenCode · Self-contained subset, API-dependent tasks excluded · Mean of five runs27.20*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- OpenCode*
- Tasks
- 78*
- Runs
- 5*
- Statistic
- Mean of five runs*
- Task set
- Self-contained subset, API-dependent tasks excluded*
- SuperGPQA · version unreported · reported score65.60*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- SWE-bench Multilingual · version unreported · reported score · Internal agent scaffold: bash and file editing · Public benchmark69.30*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- Internal agent scaffold: bash and file editing*
- Temperature
- 1.0*
- Top p
- 0.95*
- Context length
- 200K tokens*
- Task set
- Public benchmark*
- SWE-bench Pro · version unreported · reported score · Internal agent scaffold: bash and file editing · Corrected public SWE-bench Pro task set51.20*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- Internal agent scaffold: bash and file editing*
- Temperature
- 1.0*
- Top p
- 0.95*
- Context length
- 200K tokens*
- Task set
- Corrected public SWE-bench Pro task set*
- SWE-bench Verified · version unreported · reported score · Internal agent scaffold: bash and file editing · Public benchmark75*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- Internal agent scaffold: bash and file editing*
- Temperature
- 1.0*
- Top p
- 0.95*
- Context length
- 200K tokens*
- Task set
- Public benchmark*
- TAU3-Bench · version unreported · reported score68.40*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- User model
- gpt-5.2, low reasoning effort*
- Retrieval
- Default BM25*
- Terminal-Bench 2.0 · 2.0 · reported score · Harbor/Terminus-2 · Mean of five runs41.60*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- Harbor/Terminus-2*
- Timeout
- 3 hours*
- Hardware
- 32 CPU / 48 GB RAM*
- Runs
- 5*
- Statistic
- Mean of five runs*
- Temperature
- 1.0*
- Top p
- 0.95*
- Top k
- 20*
- Maximum output
- 80K tokens*
- Context length
- 256K tokens*
- Tool Decathlon · version unreported · reported score31.50*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- VideoMME (w sub.) · version unreported · reported score87*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Variant
- Subtitles enabled*
- VideoMME (w/o sub.) · version unreported · reported score82.80*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Variant
- Subtitles disabled*
- VideoMMMU · version unreported · reported score82.30*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- VITA-Bench · version unreported · reported score · Average subdomain score41.80*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Judge
- claude-4-sonnet*
- Statistic
- Average subdomain score*
- WideSearch · version unreported · reported score66.40*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- ZEROBench_sub · version unreported · reported score36.20*Platform unreported Compare results
Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
Recorded results
Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.
Arena text preference / chinesehuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,477.76 rating
Confidence interval: 1,464.65 – 1,490.87
2063 votes
- Category
- chinese
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,426.35 rating
Confidence interval: 1,419.27 – 1,433.44
7929 votes
- Category
- coding
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,362.44 rating
Confidence interval: 1,353.08 – 1,371.79
4299 votes
- Category
- creative_writing
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,424.88 rating
Confidence interval: 1,419.05 – 1,430.70
12734 votes
- Category
- english
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,391.86 rating
Confidence interval: 1,386.08 – 1,397.64
20895 votes
- Category
- exclude_ties
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / experthuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,427.34 rating
Confidence interval: 1,415.76 – 1,438.93
2646 votes
- Category
- expert
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,409.55 rating
Confidence interval: 1,388.44 – 1,430.66
847 votes
- Category
- french
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,393.28 rating
Confidence interval: 1,366.12 – 1,420.44
450 votes
- Category
- german
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,414.38 rating
Confidence interval: 1,409.17 – 1,419.58
18072 votes
- Category
- hard_prompts
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,428.39 rating
Confidence interval: 1,421.41 – 1,435.38
8331 votes
- Category
- hard_prompts_english
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,398.75 rating
Confidence interval: 1,390.58 – 1,406.91
5603 votes
- Category
- industry_business_and_management_and_financial_operations
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,363.02 rating
Confidence interval: 1,354.63 – 1,371.42
5381 votes
- Category
- industry_entertainment_and_sports_and_media
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,405.61 rating
Confidence interval: 1,392.81 – 1,418.42
2187 votes
- Category
- industry_legal_and_government
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,436.46 rating
Confidence interval: 1,427.78 – 1,445.14
4783 votes
- Category
- industry_life_and_physical_and_social_science
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,429.18 rating
Confidence interval: 1,414.49 – 1,443.86
1671 votes
- Category
- industry_mathematical
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,425.24 rating
Confidence interval: 1,412.01 – 1,438.47
2033 votes
- Category
- industry_medicine_and_healthcare
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,431.22 rating
Confidence interval: 1,425.10 – 1,437.34
11348 votes
- Category
- industry_software_and_it_services
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,380.50 rating
Confidence interval: 1,372.84 – 1,388.16
6493 votes
- Category
- industry_writing_and_literature_and_language
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,393.00 rating
Confidence interval: 1,386.40 – 1,399.60
9231 votes
- Category
- instruction_following
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / japanesehuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,345.58 rating
Confidence interval: 1,305.71 – 1,385.46
226 votes
- Category
- japanese
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / koreanhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,358.81 rating
Confidence interval: 1,330.05 – 1,387.57
446 votes
- Category
- korean
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,412.92 rating
Confidence interval: 1,406.55 – 1,419.29
11046 votes
- Category
- longer_query
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,429.01 rating
Confidence interval: 1,415.01 – 1,443.00
1746 votes
- Category
- math
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,410.15 rating
Confidence interval: 1,401.31 – 1,418.98
4887 votes
- Category
- multi_turn
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,389.75 rating
Confidence interval: 1,384.45 – 1,395.05
15913 votes
- Category
- non_english
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,408.99 rating
Confidence interval: 1,404.72 – 1,413.26
28647 votes
- Category
- overall
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / polishhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,374.56 rating
Confidence interval: 1,349.73 – 1,399.39
537 votes
- Category
- polish
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,389.57 rating
Confidence interval: 1,379.25 – 1,399.89
3313 votes
- Category
- russian
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / spanishhuman preference · 2026-10-28 · rating
Qwen3.5 27B
1,407.48 rating
Confidence interval: 1,388.12 – 1,426.83
941 votes
- Category
- spanish
- Evaluated configuration
- Qwen3.5-27B
- Evaluation type
- human preference
Arena text preference / chinesehuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,477.99 rating
Confidence interval: 1,464.86 – 1,491.12
2058 votes
- Category
- chinese
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / codinghuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,426.36 rating
Confidence interval: 1,419.28 – 1,433.45
7939 votes
- Category
- coding
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / creative_writinghuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,362.11 rating
Confidence interval: 1,352.76 – 1,371.47
4298 votes
- Category
- creative_writing
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / englishhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,424.87 rating
Confidence interval: 1,419.05 – 1,430.69
12753 votes
- Category
- english
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,391.71 rating
Confidence interval: 1,385.93 – 1,397.49
20910 votes
- Category
- exclude_ties
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / experthuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,427.22 rating
Confidence interval: 1,415.65 – 1,438.79
2654 votes
- Category
- expert
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / frenchhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,408.46 rating
Confidence interval: 1,387.37 – 1,429.55
848 votes
- Category
- french
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / germanhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,392.52 rating
Confidence interval: 1,365.34 – 1,419.70
451 votes
- Category
- german
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / hard_promptshuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,414.34 rating
Confidence interval: 1,409.13 – 1,419.54
18084 votes
- Category
- hard_prompts
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,428.43 rating
Confidence interval: 1,421.45 – 1,435.42
8348 votes
- Category
- hard_prompts_english
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,398.57 rating
Confidence interval: 1,390.41 – 1,406.74
5610 votes
- Category
- industry_business_and_management_and_financial_operations
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,363.01 rating
Confidence interval: 1,354.62 – 1,371.41
5387 votes
- Category
- industry_entertainment_and_sports_and_media
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,405.41 rating
Confidence interval: 1,392.61 – 1,418.21
2192 votes
- Category
- industry_legal_and_government
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,436.69 rating
Confidence interval: 1,428.01 – 1,445.37
4778 votes
- Category
- industry_life_and_physical_and_social_science
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,429.02 rating
Confidence interval: 1,414.33 – 1,443.70
1672 votes
- Category
- industry_mathematical
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,425.08 rating
Confidence interval: 1,411.86 – 1,438.29
2038 votes
- Category
- industry_medicine_and_healthcare
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,431.14 rating
Confidence interval: 1,425.02 – 1,437.27
11357 votes
- Category
- industry_software_and_it_services
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,380.41 rating
Confidence interval: 1,372.74 – 1,388.07
6497 votes
- Category
- industry_writing_and_literature_and_language
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,392.69 rating
Confidence interval: 1,386.08 – 1,399.30
9229 votes
- Category
- instruction_following
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / japanesehuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,345.65 rating
Confidence interval: 1,305.71 – 1,385.59
226 votes
- Category
- japanese
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / koreanhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,358.96 rating
Confidence interval: 1,330.20 – 1,387.72
448 votes
- Category
- korean
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / longer_queryhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,412.89 rating
Confidence interval: 1,406.52 – 1,419.26
11050 votes
- Category
- longer_query
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / mathhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,429.10 rating
Confidence interval: 1,415.09 – 1,443.10
1746 votes
- Category
- math
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,410.03 rating
Confidence interval: 1,401.19 – 1,418.87
4893 votes
- Category
- multi_turn
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,389.62 rating
Confidence interval: 1,384.32 – 1,394.93
15918 votes
- Category
- non_english
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,408.90 rating
Confidence interval: 1,404.63 – 1,413.17
28671 votes
- Category
- overall
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / polishhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,374.48 rating
Confidence interval: 1,349.64 – 1,399.31
537 votes
- Category
- polish
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / russianhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,389.79 rating
Confidence interval: 1,379.47 – 1,400.11
3317 votes
- Category
- russian
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Arena text preference / spanishhuman preference · 2026-10-15 · rating
Qwen3.5 27B
1,407.09 rating
Confidence interval: 1,387.71 – 1,426.47
939 votes
- Category
- spanish
- Evaluated configuration
- qwen3.5-27b
- Evaluation type
- human preference
Run it locally
API cost
Standard text-token estimate. Excludes taxes, caching, batch discounts, reasoning and multimodal surcharges. Provider limits and regional availability may differ.
Qwen3.5 27B
Qwen3.5 27B / Pricing tiers
| Provider / scope | Rate | Currency | Amount / unit |
|---|---|---|---|
| DeepInfraQwen/Qwen3.5-27BStandard | Input | USD | 0.26 / 1M tokens |
| Output | USD | 2.6 / 1M tokens | |
| SiliconFlowQwen/Qwen3.5-27BStandard 1314 | Input | CNY | 1.8 / 1M tokens |
| Output | CNY | 14.4 / 1M tokens | |
| SiliconFlowQwen/Qwen3.5-27BStandard 输入 [0, 128k) Token tier: 0 – 128,000 | Input | CNY | 0.6 / 1M tokens |
| Output | CNY | 4.8 / 1M tokens | |
| SiliconFlowQwen/Qwen3.5-27BStandard 输入 [0,128k) Token tier: 0 – 128,000 | Input | CNY | 0.6 / 1M tokens |
| Output | CNY | 4.8 / 1M tokens | |
| SiliconFlowQwen/Qwen3.5-27BStandard 输入 [128k, +∞) Token tier: 128,000 – No stated upper limit | Input | CNY | 1.8 / 1M tokens |
| Output | CNY | 14.4 / 1M tokens | |
| SiliconFlowQwen/Qwen3.5-27BStandard 输入 [128k,256k) Token tier: 128,000 – 256,000 | Input | CNY | 1.8 / 1M tokens |
| Output | CNY | 14.4 / 1M tokens |
Capabilities
- Context window
- 262,144 tokens
- Maximum output
- 65,536 tokens
- Reasoning
- Yes
- Tool calling
- Yes
- Open weights
- Yes
- Checkpoint tensor payload
- 51.75 GiB
Input modalities
- text
- image
- video
Output modalities
- text
Processing & platform
- Architecture
- qwen3_5
System configuration
Detailed specifications
Specifications · 51 fields
- HallusionBench
- suite
- HallusionBench
- workload
- reported score
- score
- 70.0*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- NL2Repo
- suite
- NL2Repo
- workload
- reported score · Claude Code
- score
- 27.3*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- Claude Code*
- Variant
- Maximum 900 turns*
- Temperature
- 1.0*
- Top p
- 0.95*
- QwenWebBench
- suite
- QwenWebBench
- workload
- reported score · Internal
- score
- 1,068.0*
- unit
- Elo
Scope and conditions
- Test conditions
- Metric
- BT/Elo rating*
- Checkpoint version
- Not specified*
- Languages
- English and Chinese*
- Categories
- Seven frontend application categories*
- Judge
- Multimodal automatic judge*
- Task set
- Internal*
- OmniDocBench1.5
- suite
- OmniDocBench1.5
- version
- 1.5
- workload
- reported score
- score
- 88.9*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- CharXiv(RQ)
- suite
- CharXiv(RQ)
- workload
- reported score
- score
- 79.5*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- TAU3-Bench
- suite
- TAU3-Bench
- workload
- reported score
- score
- 68.4*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- User model
- gpt-5.2, low reasoning effort*
- Retrieval
- Default BM25*
- CC-OCR
- suite
- CC-OCR
- workload
- reported score
- score
- 81.0*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- VITA-Bench
- suite
- VITA-Bench
- workload
- reported score · Average subdomain score
- score
- 41.8*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Judge
- claude-4-sonnet*
- Statistic
- Average subdomain score*
- AI2D TEST
- suite
- AI2D_TEST
- workload
- reported score
- score
- 92.9*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- DeepPlanning
- suite
- DeepPlanning
- workload
- reported score
- score
- 22.6*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Tool Decathlon
- suite
- Tool Decathlon
- workload
- reported score
- score
- 31.5*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- RefCOCO(avg)
- suite
- RefCOCO(avg)
- workload
- reported score
- score
- 90.9*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MCPMark
- suite
- MCPMark
- workload
- reported score · GitHub MCP v0.30.3
- score
- 36.3*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- GitHub MCP v0.30.3*
- Truncation
- Playwright responses at 32K tokens*
- MCP-Atlas
- suite
- MCP-Atlas
- workload
- reported score · Public set
- score
- 68.4*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Judge
- gemini-2.5-pro*
- Task set
- Public set*
- ODInW13
- suite
- ODInW13
- workload
- reported score
- score
- 41.1*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- EmbSpatialBench
- suite
- EmbSpatialBench
- workload
- reported score
- score
- 84.5*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- WideSearch
- suite
- WideSearch
- workload
- reported score
- score
- 66.4*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MMMU
- suite
- MMMU
- workload
- reported score
- score
- 82.3*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- SWE-bench Verified
- suite
- SWE-bench Verified
- workload
- reported score · Internal agent scaffold: bash and file editing · Public benchmark
- score
- 75.0*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- Internal agent scaffold: bash and file editing*
- Temperature
- 1.0*
- Top p
- 0.95*
- Context length
- 200K tokens*
- Task set
- Public benchmark*
- RefSpatialBench
- suite
- RefSpatialBench
- workload
- reported score
- score
- 67.7*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MMLU-Pro
- suite
- MMLU-Pro
- workload
- reported score
- score
- 86.1*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MMLU-Redux
- suite
- MMLU-Redux
- workload
- reported score
- score
- 93.2*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- VideoMME (w sub.)
- suite
- VideoMME (w sub.)
- workload
- reported score
- score
- 87.0*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Variant
- Subtitles enabled*
- VideoMME (w/o sub.)
- suite
- VideoMME (w/o sub.)
- workload
- reported score
- score
- 82.8*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Variant
- Subtitles disabled*
- SuperGPQA
- suite
- SuperGPQA
- workload
- reported score
- score
- 65.6*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- VideoMMMU
- suite
- VideoMMMU
- workload
- reported score
- score
- 82.3*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- C-Eval
- suite
- C-Eval
- workload
- reported score
- score
- 90.5*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MLVU
- suite
- MLVU
- workload
- reported score
- score
- 85.9*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- GPQA
- suite
- GPQA
- workload
- reported score
- score
- 85.5*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MVBench
- suite
- MVBench
- workload
- reported score
- score
- 74.6*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- HLE
- suite
- HLE
- workload
- reported score
- score
- 24.3*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- LVBench
- suite
- LVBench
- workload
- reported score
- score
- 73.6*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- LiveCodeBench v6
- suite
- LiveCodeBench v6
- version
- 6
- workload
- reported score
- score
- 80.7*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- HMMT Feb 25
- suite
- HMMT Feb 25
- workload
- reported score
- score
- 92.0*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- SWE-bench Multilingual
- suite
- SWE-bench Multilingual
- workload
- reported score · Internal agent scaffold: bash and file editing · Public benchmark
- score
- 69.3*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- Internal agent scaffold: bash and file editing*
- Temperature
- 1.0*
- Top p
- 0.95*
- Context length
- 200K tokens*
- Task set
- Public benchmark*
- MMMU-Pro
- suite
- MMMU-Pro
- workload
- reported score
- score
- 75.0*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- HMMT Nov 25
- suite
- HMMT Nov 25
- workload
- reported score
- score
- 89.8*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- HMMT Feb 26
- suite
- HMMT Feb 26
- workload
- reported score
- score
- 84.3*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- IMOAnswerBench
- suite
- IMOAnswerBench
- workload
- reported score
- score
- 79.9*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- AIME26
- suite
- AIME26
- workload
- reported score · Full AIME 2026 I and II
- score
- 92.6*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Task set
- Full AIME 2026 I and II*
- Mathvista(mini)
- suite
- Mathvista(mini)
- workload
- reported score
- score
- 87.8*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- SWE-bench Pro
- suite
- SWE-bench Pro
- workload
- reported score · Internal agent scaffold: bash and file editing · Corrected public SWE-bench Pro task set
- score
- 51.2*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- Internal agent scaffold: bash and file editing*
- Temperature
- 1.0*
- Top p
- 0.95*
- Context length
- 200K tokens*
- Task set
- Corrected public SWE-bench Pro task set*
- Terminal-Bench 2.0
- suite
- Terminal-Bench 2.0
- version
- 2.0
- workload
- reported score · Harbor/Terminus-2 · Mean of five runs
- score
- 41.6*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- Harbor/Terminus-2*
- Timeout
- 3 hours*
- Hardware
- 32 CPU / 48 GB RAM*
- Runs
- 5*
- Statistic
- Mean of five runs*
- Temperature
- 1.0*
- Top p
- 0.95*
- Top k
- 20*
- Maximum output
- 80K tokens*
- Context length
- 256K tokens*
- ZEROBench sub
- suite
- ZEROBench_sub
- workload
- reported score
- score
- 36.2*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Claw-Eval Avg
- suite
- Claw-Eval Avg
- workload
- reported score
- score
- 64.3*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Claw-Eval Pass^3
- suite
- Claw-Eval Pass^3
- workload
- reported score
- score
- 46.2*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- RealWorldQA
- suite
- RealWorldQA
- workload
- reported score
- score
- 83.7*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- MMBench EN-DEV-v1.1
- suite
- MMBench EN-DEV-v1.1
- version
- 1.1
- workload
- reported score
- score
- 92.6*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- SkillsBench Avg5
- suite
- SkillsBench Avg5
- workload
- reported score · OpenCode · Self-contained subset, API-dependent tasks excluded · Mean of five runs
- score
- 27.2*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Harness
- OpenCode*
- Tasks
- 78*
- Runs
- 5*
- Statistic
- Mean of five runs*
- Task set
- Self-contained subset, API-dependent tasks excluded*
- QwenClawBench
- suite
- QwenClawBench
- workload
- reported score · Internal real-user distribution
- score
- 52.2*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
- Task set
- Internal real-user distribution*
- Temperature
- 0.6*
- Context length
- 256K tokens*
- SimpleVQA
- suite
- SimpleVQA
- workload
- reported score
- score
- 56.0*
Scope and conditions
- Test conditions
- Metric
- reported score*
- Checkpoint version
- Not specified*
0 selected Compare 0
* Provisional data