Qwen3-235B-A22B-Instruct-2507

CurrentVerified · Jun 27, 2026

Largest open-weight Qwen3 MoE flagship (235B total / 22B active), Apache 2.0. QwenLM/Qwen3 GitHub README: '2025.07.21: We released the updated version of Qwen3-235B-A22B non-thinking mode, named Qwen3-235B-A22B-Instruct-2507, featuring significant enhancements over the previous version and supportin

profile normalized against the 70-model field

Context window· 262K of 10M3%

Max output· 16K of 384K4%

Output speed—

Affordability—

Capability breadth· 1 of 119%

Capability switches · 1 of 11

Reasoning mode

Tool / function use

Streaming

JSON mode

Structured outputs

Prompt caching

Fine-tuning

Web search

Code execution

Vision input

Audio input

Specifications

Every value carries a primary source and a verification date.

Capacity

Context window

262K

Max output

16K

Capabilities

Reasoning mode

Tool / function use

Yes

Vision input

API

API model ID

Qwen/Qwen3-235B-A22B-Instruct-2507

General

Knowledge cutoff

256K natively and extendable up to 1M tokens

Release date

July 21, 2025

Benchmarks

Sourced evaluation scores, each verified against its primary source.

MMLU-Pro

| MMLU-Pro | 81.2 | 79.8 | **86.6** | 81.1 | 75.2 | 83.0 |

83 accuracy %Verified

MMLU-Redux

| MMLU-Redux | 90.4 | 91.3 | **94.2** | 92.7 | 89.2 | 93.1 |

93.1 accuracy %Verified

GPQA Diamond

| GPQA | 68.4 | 66.9 | 74.9 | 75.1 | 62.9 | **77.5** |

77.5 accuracy %Verified

SuperGPQA

| SuperGPQA | 57.3 | 51.0 | 56.5 | 57.2 | 48.2 | **62.6** |

62.6 accuracy %Verified

SimpleQA

| SimpleQA | 27.2 | 40.3 | 22.8 | 31.0 | 12.2 | **54.3** |

54.3 accuracy %Verified

AIME25

| AIME25 | 46.6 | 26.7 | 33.9 | 49.5 | 24.7 | **70.3** |

70.3 accuracy %Verified

HMMT25

| HMMT25 | 27.5 | 7.9 | 15.9 | 38.8 | 10.0 | **55.4** |

55.4 accuracy %Verified

ARC-AGI

| ARC-AGI | 9.0 | 8.8 | 30.3 | 13.3 | 4.3 | **41.8** |

41.8 accuracy %Verified

ZebraLogic

| ZebraLogic | 83.4 | 52.6 | - | 89.0 | 37.7 | **95.0** |

95 accuracy %Verified

LiveBench 20241125

| LiveBench 20241125 | 66.9 | 63.7 | 74.6 | **76.4** | 62.5 | 75.4 |

75.4 accuracy %Verified

LiveCodeBench v6 (25.02-25.05)

| LiveCodeBench v6 (25.02-25.05) | 45.2 | 35.8 | 44.6 | 48.9 | 32.9 | **51.8** |

51.8 accuracy %Verified

MultiPL-E

| MultiPL-E | 82.2 | 82.7 | **88.5** | 85.7 | 79.3 | 87.9 |

87.9 accuracy %Verified

Aider-Polyglot

| Aider-Polyglot | 55.1 | 45.3 | **70.7** | 59.0 | 59.6 | 57.3 |

57.3 accuracy %Verified

IFEval

| IFEval | 82.3 | 83.9 | 87.4 | **89.8** | 83.2 | 88.7 |

88.7 accuracy %Verified

Arena-Hard v2

| Arena-Hard v2* | 45.6 | 61.9 | 51.5 | 66.1 | 52.0 | **79.2** |

79.2 win rate %Verified

Creative Writing v3

| Creative Writing v3 | 81.6 | 84.9 | 83.8 | **88.1** | 80.4 | 87.5 |

87.5 scoreVerified

WritingBench

| WritingBench | 74.5 | 75.5 | 79.2 | **86.2** | 77.0 | 85.2 |

85.2 scoreVerified

BFCL-v3

| BFCL-v3 | 64.7 | 66.5 | 60.1 | 65.2 | 68.0 | **70.9** |

70.9 accuracy %Verified

TAU2-Retail

| TAU2-Retail | 71.1 | 66.7# | **75.5** | 70.6 | 64.9 | 74.6 |

74.6 accuracy %Verified

MultiIF

| MultiIF | 66.5 | 70.4 | - | 76.2 | 70.2 | **77.5** |

77.5 accuracy %Verified

MMLU-ProX

| MMLU-ProX | 75.8 | 76.2 | - | 74.5 | 73.2 | **79.4** |

79.4 accuracy %Verified

PolyMATH

| PolyMATH | 32.2 | 25.5 | 30.0 | 44.8 | 27.0 | **50.2** |

50.2 accuracy %Verified

Loading…

Qwen3-235B-A22B-Instruct-2507

CurrentVerified · Jun 27, 2026

Specifications

Every value carries a primary source and a verification date.

Capacity

Context window

262K

Max output

16K

Capabilities

Reasoning mode

Tool / function use

Yes

Vision input

API

API model ID

Qwen/Qwen3-235B-A22B-Instruct-2507

General

Knowledge cutoff

256K natively and extendable up to 1M tokens

Release date

July 21, 2025

Benchmarks

Sourced evaluation scores, each verified against its primary source.

MMLU-Pro

| MMLU-Pro | 81.2 | 79.8 | **86.6** | 81.1 | 75.2 | 83.0 |

83 accuracy %Verified

MMLU-Redux

| MMLU-Redux | 90.4 | 91.3 | **94.2** | 92.7 | 89.2 | 93.1 |

93.1 accuracy %Verified

GPQA Diamond

| GPQA | 68.4 | 66.9 | 74.9 | 75.1 | 62.9 | **77.5** |

77.5 accuracy %Verified

SuperGPQA

| SuperGPQA | 57.3 | 51.0 | 56.5 | 57.2 | 48.2 | **62.6** |

62.6 accuracy %Verified

SimpleQA

| SimpleQA | 27.2 | 40.3 | 22.8 | 31.0 | 12.2 | **54.3** |

54.3 accuracy %Verified

AIME25

| AIME25 | 46.6 | 26.7 | 33.9 | 49.5 | 24.7 | **70.3** |

70.3 accuracy %Verified

HMMT25

| HMMT25 | 27.5 | 7.9 | 15.9 | 38.8 | 10.0 | **55.4** |

55.4 accuracy %Verified

ARC-AGI

| ARC-AGI | 9.0 | 8.8 | 30.3 | 13.3 | 4.3 | **41.8** |

41.8 accuracy %Verified

ZebraLogic

| ZebraLogic | 83.4 | 52.6 | - | 89.0 | 37.7 | **95.0** |

95 accuracy %Verified

LiveBench 20241125

| LiveBench 20241125 | 66.9 | 63.7 | 74.6 | **76.4** | 62.5 | 75.4 |

75.4 accuracy %Verified

LiveCodeBench v6 (25.02-25.05)

| LiveCodeBench v6 (25.02-25.05) | 45.2 | 35.8 | 44.6 | 48.9 | 32.9 | **51.8** |

51.8 accuracy %Verified

MultiPL-E

| MultiPL-E | 82.2 | 82.7 | **88.5** | 85.7 | 79.3 | 87.9 |

87.9 accuracy %Verified

Aider-Polyglot

| Aider-Polyglot | 55.1 | 45.3 | **70.7** | 59.0 | 59.6 | 57.3 |

57.3 accuracy %Verified

IFEval

| IFEval | 82.3 | 83.9 | 87.4 | **89.8** | 83.2 | 88.7 |

88.7 accuracy %Verified

Arena-Hard v2

| Arena-Hard v2* | 45.6 | 61.9 | 51.5 | 66.1 | 52.0 | **79.2** |

79.2 win rate %Verified

Creative Writing v3

| Creative Writing v3 | 81.6 | 84.9 | 83.8 | **88.1** | 80.4 | 87.5 |

87.5 scoreVerified

WritingBench

| WritingBench | 74.5 | 75.5 | 79.2 | **86.2** | 77.0 | 85.2 |

85.2 scoreVerified

BFCL-v3

| BFCL-v3 | 64.7 | 66.5 | 60.1 | 65.2 | 68.0 | **70.9** |

70.9 accuracy %Verified

TAU2-Retail

| TAU2-Retail | 71.1 | 66.7# | **75.5** | 70.6 | 64.9 | 74.6 |

74.6 accuracy %Verified

MultiIF

| MultiIF | 66.5 | 70.4 | - | 76.2 | 70.2 | **77.5** |

77.5 accuracy %Verified

MMLU-ProX

| MMLU-ProX | 75.8 | 76.2 | - | 74.5 | 73.2 | **79.4** |

79.4 accuracy %Verified

PolyMATH

| PolyMATH | 32.2 | 25.5 | 30.0 | 44.8 | 27.0 | **50.2** |

50.2 accuracy %Verified