Skip to content
llmwaves

Qwen3 235B A22B Thinking 2507

Qwen · released Jul 25, 2025

Open weightsReasoningTool use

Qwen3-235B-A22B-Thinking-2507 is a high-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks. It activates 22B of its 235B parameters per forward pass and natively supports up to 262,144...

Specification

Context window
128K
Max output
16K
Knowledge cutoff
Jun 30, 2025
Parameters
235.1B
Licence
apache-2.0
Serving providers
1
Moderated
No
Uptime
100.0%

Intelligence

—

Not benchmarked

Coding

—

Coding Index

Agentic

—

Agentic Index

Output speed

6 t/s

Median across providers

Latency

13.73s

Time to first token

Cost per task

$0.30

Estimated

Benchmarks

Where the score comes from

The Intelligence Index is a composite. These are the underlying evaluations this model was actually measured on.

Evaluation scores

This model has not been put through the benchmark suite. Its speed, latency and pricing figures below are still measured.

Against its peers

Intelligence Index · this model highlighted, nearest peers in grey

Peers are the models sitting closest on the Intelligence Index — the set you would realistically choose between.

View as table
ModelIntelligence
Qwen3.8 Max58.1
Claude Opus 5.557.6
Claude Sonnet 5.556.0
Claude Fable 5.153.4
GPT-6 Astra52.7
Gemini 4 Argon52.6
Muse Wonder52.4

Percentile among all indexed models

Arena Elo18th

Pricing

What it costs to run

List prices per million tokens, plus what one representative task works out to.

List price

Input / 1M tokens
$0.45
Output / 1M tokens
$3.5
Cached input / 1M
Not offered
Blended 3:1
$1.21

One task, estimated

$0.30

Input tokens
50,000
Output tokens
80,000
Profile
Reasoning

Estimated from list pricing: 50K input tokens plus 80K output tokens for reasoning models (25K for non-reasoning).

Serving providers

Speed and latency figures are medians across these providers, so a widely-served model reports a blend rather than any single endpoint.

Arena

Head-to-head generation quality

Elo from pairwise judgements, broken out by the kind of thing the model was asked to build.

Overall Elo
1,054
Win rate
42.0%
Strongest at
Websites
Tournaments
4,938

Elo by category

Dot position on a 925–1,075 scale · Elo has no meaningful zero

Agent categories (full-stack apps, mobile apps) are only scored for models tested in agent harnesses.

View as table
CategoryEloWin rate
Websites105442.0%
3D scenes99840.4%
Game development96034.2%
UI components94634.0%