qwen Open weights

Qwen3 235B A22B Instruct 2507 pricing & benchmarks

Released Jul 21, 2025 · served by 11 providers

Input / 1M tokens$0.090
Output / 1M tokens$0.550
Context window262K
Cached input
1M in + 300K out$0.26

What Qwen3 235B A22B Instruct 2507 is

Qwen3-235B-A22B-Instruct-2507 is a multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following,...

Providers and prices for Qwen3 235B A22B Instruct 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
DeepInfracheapest $0.090 $0.550 262K fp8 97.9%
Novita $0.090 $0.580 131K fp8 97.8%
Alibaba $0.149 $0.598 131K 100.0%
Nebius $0.200 $0.600 262K fp8 92.7%
Venice $0.150 $0.750 128K fp8 96.8%
Parasail $0.140 $0.800 131K fp8 99.6%
Friendli $0.200 $0.800 262K 94.8%
Crusoe $0.220 $0.800 262K bf16 99.4%
StreamLake $0.210 $0.840 128K 99.8%
AtlasCloud $0.200 $0.880 131K fp8 97.6%
Google $0.250 $1.00 262K 99.0%

Qwen3 235B A22B Instruct 2507 benchmark results

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

CategoryArenaRankEloWin rate
dataviz models #85 1087 47.7%
3d models #89 1051 41.1%
codecategories models #89 1069 42.7%
website models #91 1083 43.7%
uicomponent models #93 1003 38.8%
gamedev models #101 1008 35.2%

FAQ

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

$0.090 per 1M input tokens and $0.550 per 1M output tokens. A typical workload of 1M input + 300K output tokens costs about $0.26.

Is Qwen3 235B A22B Instruct 2507 free?

No. It is a paid model starting at $0.090 per 1M input tokens, though some providers offer trial credits.

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

DeepInfra at $0.090 input / $0.550 output per 1M tokens (fp8 quantization).

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

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