Skip to content
aiqbench

Alibaba · model passport

Qwen: Qwen3 235B A22B Thinking 2507

qwen3-235b-a22b-thinking-2507
  • Open weights
  • Apache 2.0 licence
  • 235.1B params
  • 128,000 context
  • Released 2025-07-01(from model name)
8/27axes covered
Overall · LMArena
1,400#150 of 396
Best percentile
67.0AIME (math)
Cohort standing
56.9median 51.6

What the data says

Benchmarked on 8 of 27 tracked axes. Every figure below is the model's own published result on that board.

Uneven: 24.3 percentile points between its best and worst axis — 67 pct on AIME (math) versus 42.7 pct on SimpleQA Verified.

  • AIME (math)67 pct86.7%Verify →
  • Overall62.4 pct1,400Verify →
  • Instruction following61.9 pct1,385Verify →
  • GPQA Diamond61.1 pct80.1%Verify →
  • Coding61.1 pct1,442Verify →
  • Hard prompts61.1 pct1,417Verify →
  • LiveBench58.4 pct57.8%Verify →
  • SimpleQA Verified42.7 pct40.4%Verify →
  • WebDev ArenaNot benchmarked
  • Vision ArenaNot benchmarked
  • Factuality ArenaNot benchmarked
  • Search ArenaNot benchmarked
  • Document ArenaNot benchmarked
  • Agent ArenaNot benchmarked
  • FrontierMathNot benchmarked
  • ARC-AGI-2Not benchmarked
  • Humanity's Last ExamNot benchmarked
  • SimpleBenchNot benchmarked
  • BFCL v4 (function calling)Not benchmarked
  • SWE-bench Verified (live board)Not benchmarked
  • Terminal-Bench 4.0Not benchmarked
  • MMLU-Pro (Artificial Analysis)Not benchmarked
  • LiveCodeBench (Artificial Analysis)Not benchmarked
  • IFBench (Artificial Analysis)Not benchmarked
  • HHEM Hallucination (Vectara)Not benchmarked
  • SWE-bench VerifiedNot benchmarked
  • Terminal-BenchNot benchmarked

Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.

How to run it

Run it yourself

RepositoryQwen/Qwen3-235B-A22B-Thinking-2507
Parameters235.1B
ArchitectureQwen3MoeForCausalLM
LicenceApache 2.0

Transformers

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "Qwen/Qwen3-235B-A22B-Thinking-2507"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

Ollama

We don't sync Ollama tags, so we won't print a command we can't verify — a wrong one would pull the wrong model. Search the Ollama library →

Unsloth fine-tuning recipes are model-specific and we don't derive them here. Unsloth docs →

Use it hosted

API model IDqwen/qwen3-235b-a22b-thinking-2507
Context length128,000 tokens
Max output16,384 tokens
Input price$0.45 per 1M tokens
Output price$3.50 per 1M tokens

API parameters this endpoint supports

Frequency penaltyInclude reasoningMax tokensPresence penaltyReasoningStop sequencesTemperatureTool choiceTool callingTop-kTop-p

Example request

curl https://openrouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen/qwen3-235b-a22b-thinking-2507","messages":[{"role":"user","content":"Hello"}]}'

View on OpenRouter →

Identifiers, limits and prices are synced from the upstream source shown above. Any field we haven't synced is left out entirely rather than guessed.

  • qwen3-235b-a22b-thinking-2507 stands at 56.9 (raw mean 59.5) — measured on 8 of 27 axes.

    standingtemplateVerify →

  • qwen3-235b-a22b-thinking-2507 stands at 56.9 versus a cohort median of 51.6 across 144 models released within 180 days — above the peer median.

    cohorttemplateVerify →

Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →