Inference
EN
Alibaba

Qwen3 235B-A22B

Qwen3 235B-A22B is a standard model from Alibaba. 14 benchmarks count toward its index, in 5 categories.

  • Open weights
  • Standard
  • Input: text
  • Released 2025-04
  • Retires 9 Oct 2026
RankedEnough results for an index.
IndexAll categories, out of 100.38.1±8.8Median 51.0
CoverageIndex weight that has results.55%
SpeedOutput tokens per second.44/sMedian 67/s
First tokenSeconds to the first output token.0.7 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.455Median $0.60
Output / 1MUS dollars per 1M output tokens.$1.82Median $2.55
ContextMaximum tokens in one request.131KMedian 262K
EloLMArena rating and rank.1366±5 #194Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
qwen/qwen3-235b-a22b
Max outputMaximum output tokens per request.
8K
Knowledge cutoffLast date of training data.
31 Mar 2025
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
No
Cache read / 1MUS dollars per 1M cached tokens.
—
WeightsLink to the published weights.
Hugging Face (opens in a new tab) ↗
RetiresDate OpenRouter removes the model.
9 Oct 2026

Run locally

235.09B parameters
16-bit
494.7 GB
8-bit
263.3 GB
4-bit
150.6 GB
Weights
Hugging Face (opens in a new tab) ↗

Memory need with a short context. The 4-bit copy needs more than 128 GB. Compare local models

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.None counted——
CodingCode writing and repair.2 counted44.0167 / 209
ReasoningLogic problems and puzzles.None counted——
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.5 counted41.1189 / 268
MultilingualTasks in many languages.1 counted42.0—
InstructionTasks with strict prompt rules.1 counted30.7—
MathMath problems.5 counted32.2140 / 168

Results

14 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
MATH 500Math94.6% ±1.08 / 59
MATH level 5Math68.9% ±0.934 / 88
AIMEMath84.0% ±0.840 / 95
GPQA diamond · Epoch AIKnowledge70.7% ±2.7113 / 217
LiveCodeBenchCoding70.6% ±1.186 / 142
MMLU ProKnowledge81.2% ±0.482 / 135
MMLU-ReduxKnowledge89.2%12 / 19
Instruction-Following EvalInstruction83.2%46 / 70
GPQA Diamond · Vals AIKnowledge70.2% ±2.492 / 135
MATH-500 Problem SetMath91.2%20 / 27
INCLUDEMultilingual75.6%9 / 12
Graduate-Level Google-Proof Q&AKnowledge62.9%108 / 141
C-EvalKnowledge86.1%10 / 12
American Invitational Mathematics Examination 2024Math40.1%23 / 27
MMMLUKnowledge79.8%13 / 15
American Invitational Mathematics Examination 2025Math24.7%50 / 53
IOI v1Coding0.0%61 / 61
MGSMMultilingual92.5% ±1.617 / 74
MedQAHealthcare90.6% ±0.345 / 95
FrontierMath-2025-02-28-PublicMath0.0%33 / 64
LegalBenchLegal80.2% ±0.489 / 149
TaxEval v2Finance70.6% ±0.989 / 143
Berkeley Function Calling Leaderboard v3Agentic68.0%—
LiveCodeBench v5Coding35.3%—

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseVals AINo licence statedEpoch AICC BY

History

5 releases
Index39.4 → 38.1
23 Oct 202612 Nov 2026
Output / 1M$1.82 → $1.82
23 Oct 202612 Nov 2026

Questions

How much does Qwen3 235B-A22B cost?
$0.455 per 1M input tokens, $1.82 per 1M output tokens.
How good is Qwen3 235B-A22B?
Its index is 38.1 out of 100. That is place 219 of 283 ranked models.
What is Qwen3 235B-A22B best at?
Knowledge: 41.1 out of 100, place 189 of 268.
How fast is Qwen3 235B-A22B?
44 output tokens per second, 0.7 s to the first token.
What is the context window of Qwen3 235B-A22B?
131K tokens.
Is Qwen3 235B-A22B open source?
Yes. Its weights are published.
When was Qwen3 235B-A22B released?
April 2025.

Badge

Qwen3 235B-A22B on Inference360: place 219 of 283, index 38.1
HTML
<a href="https://inference360.ai/models/qwen3-235b-a22b"><img src="https://inference360.ai/badges/qwen3-235b-a22b.svg" alt="Qwen3 235B-A22B on Inference360: place 219 of 283, index 38.1" height="20"></a>
Markdown
[![Qwen3 235B-A22B on Inference360: place 219 of 283, index 38.1](https://inference360.ai/badges/qwen3-235b-a22b.svg)](https://inference360.ai/models/qwen3-235b-a22b)

Compare with

GPT-4.138.0Nemotron 3 Nano Omni 30B A3B38.6QwQ-32B36.3Qwen3 235B 250742.3

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeNemotron 3 Ultra50.8 · Free

More from Alibaba

All models
Qwen3 235B-A22B (Reasoning)UnrankedQwen3 32BUnrankedQwen3 32B (Reasoning)UnrankedQwen3 30B A3BUnrankedQwen3 8BUnranked