qwen Open weights Reasoning

Qwen3 235B A22B Thinking 2507 pricing & benchmarks

Released Jul 25, 2025 · served by 4 providers

Input / 1M tokens$0.149
Output / 1M tokens$1.50
Context window262K
Cached input
Intelligence index19.6
Coding index22.1
Agentic index3.8
1M in + 300K out$0.60

What Qwen3 235B A22B Thinking 2507 is

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...

Providers and prices for Qwen3 235B A22B Thinking 2507

The same model costs different amounts depending on who serves it. Prices are per 1M tokens, cheapest first. Uptime and throughput are OpenRouter's rolling measurements, not vendor claims.

ProviderInputOutputContext QuantUptime 24hThroughput
Alibabacheapest $0.149 $1.50 131K 99.7%
DeepInfra $0.230 $2.30 262K fp8 99.9%
Novita $0.300 $3.00 131K fp8 94.9%
Venice $0.450 $3.50 128K fp8 98.8%

Qwen3 235B A22B Thinking 2507 benchmark results

Design Arena head-to-head results, as reported through the OpenRouter model API.

CategoryArenaRankEloWin rate
3d models #88 1056 40.7%
codecategories models #92 1063 40.9%
website models #92 1077 42.1%
uicomponent models #98 979 33.9%
dataviz models #99 976 32.3%
gamedev models #100 1014 34.3%

Cheaper models in the same class

Models scoring within 4 points of Qwen3 235B A22B Thinking 2507 on the Intelligence Index, but with a lower output price.

FAQ

How much does the Qwen3 235B A22B Thinking 2507 API cost?

$0.149 per 1M input tokens and $1.50 per 1M output tokens. A typical workload of 1M input + 300K output tokens costs about $0.60.

Is Qwen3 235B A22B Thinking 2507 free?

No. It is a paid model starting at $0.149 per 1M input tokens, though some providers offer trial credits.

Which provider is cheapest for Qwen3 235B A22B Thinking 2507?

Alibaba at $0.149 input / $1.50 output per 1M tokens.

Can I self-host Qwen3 235B A22B Thinking 2507?

Yes — weights are published as Qwen/Qwen3-235B-A22B-Thinking-2507 on Hugging Face, so you can run it on your own hardware.