BlogLLM Benchmark Scores as Published, Side by Side
Qwen3-235B-A22B: Benchmark Scores as Published
Every result the catalog holds for Qwen3-235B-A22B: 51 tests in 8 areas, from 3 pages. 30 printed by Alibaba itself, the rest by the company named beside each number. Each number is as printed; nothing is ranked or rescored.
Developer AlibabaReleased 2025-04-29Context window 32.8KInput TextOutput TextWeights Open (Apache 2.0)Data version 2026103001, as of 2026-10-30
Agentic Coding
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| SWE-bench Verified | 34.4% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking; Harness: agentic, single attempt |
| SWE-bench Multilingual | 20.9% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking; Harness: agentic, single attempt |
| Terminal-Bench (version not stated) | 6.6% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking; Harness: Terminus |
| SWE-bench Verified (agentless, single patch without test) | 39.4% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking; Harness: agentless |
Science and Math
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPQA Diamond | 81.1% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages | - |
| AIME 2025 | 81.5% | DeepSeek, DeepSeek-R1-0528 model cardchecked 2026-10-15; printed on 1 more page | - |
| HMMT February 2025 | 85.1% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages | - |
| HMMT November 2025 | 89.5% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| AIME 2024 | 85.7% | DeepSeek, DeepSeek-R1-0528 model cardchecked 2026-10-15; printed on 1 more page | - |
| MATH-500 | 91.2% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| CNMO 2024 | 48.6% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| PolyMATH | 60.1% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| PolyMath (en) | 51.9% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
Long Context
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| AA-LCR | 60% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| LongBench v2 | 54.8% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
General Knowledge and Reasoning
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Humanity's Last Exam (no tools) | 18.2% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page | Tools: none |
| MMLU-Pro | 84.4% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page | - |
| MMLU-Redux | 93.8% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page | - |
| C-Eval | 92.1% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| SuperGPQA | 64.9% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page | - |
| MMLU | 87% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| SimpleQA | 13.2% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| AutoLogi | 83.3% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| LiveBench (2024-11-25) | 67.6% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| ZebraLogic | 37.7% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
Code Generation
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| LiveCodeBench v6 | 75.1% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages | - |
| Codeforces rating | 2146 | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| Aider polyglot | 61.8% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| MultiPL-E | 78.2% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| OJBench (language not stated) | 32.7% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page | - |
| FullStackBench en | 61.1% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| FullStackBench zh | 63.1% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
Tool Use and Workflows
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| τ²-bench Telecom | 22.1% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| τ²-bench Retail | 57% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| τ²-bench Airline | 26.5% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
| τ²-bench (domain not stated) | 58.5% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| DeepPlanning | 17.1% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| VitaBench | 31.6% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| BFCL v4 | 54.8% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| AceBench | 70.5% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
Instruction Following
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| IFBench | 51.7 | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| MultiChallenge | 50.2% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page | - |
| IFEval | 87.8% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| IFEval (prompt-level strict) | 83.2% | Moonshot AI, Kimi K2 release pagechecked 2026-10-15 | Effort: non-thinking |
Multilingual
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| MMMLU | 83.4% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| MMLU-ProX | 77.9% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| WMT24++ (XCOMET-XXL) | 75.8 | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| Global PIQA | 85.7% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| INCLUDE | 81% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| MaXIFE | 83.2% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| NOVA-63 | 55.4% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
More from Alibaba
- Qwen3.8-Max
- Qwen3.7-Plus
- Qwen3.6-27B
- Qwen3.6-35B-A3B
- Qwen3.6-Plus
- Qwen3.5-122B-A10B
- Qwen3.5-27B
- Qwen3.5-35B-A3B
- Qwen3.5-397B-A17B
- Qwen3-VL-32B-Instruct
- Qwen3-VL-32B-Thinking
- Qwen3-VL-8B-Instruct
- Qwen3-VL-8B-Thinking
- Qwen3-VL-235B-A22B
- Qwen3-VL-235B-A22B-Thinking
- Qwen3-Next-80B-A3B
- Qwen3-30B-A3B-Instruct-2507
- Qwen3-30B-A3B-Thinking-2507
- Qwen3-Coder-Plus
- Qwen3-32B
- Qwen3-8B
- QwQ-32B-Preview
- Qwen-Max
- Qwen3-Max-Thinking
- Qwen3.7-Max
- Qwen3.8-2.4T-A95B
- Qwen3.8-27B
- Qwen3.8-Flash-Next
Every number here is read from the same catalog file as the benchmark table, data version 2026103001. Where one page printed the model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.