All Models
Qwen3 235b A22B Instruct 2507
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,...
Benchmarks
Available Providers (16)
| Provider | Model ID | Input Cost | Output Cost | Context | Max Output | Docs |
|---|---|---|---|---|---|---|
 ModelScope | Qwen/Qwen3-235B-A22B-Instruct-2507 | $0/MTok | $0/MTok | 262.1K | 131.1K | |
 Cortecs | qwen3-235b-a22b-instruct-2507 | $0.06/MTok | $0.41/MTok | 131K | 131K | |
 Kilo Gateway | qwen/qwen3-235b-a22b-2507 | $0.07/MTok | $0.10/MTok | 262.1K | 52.4K | |
 NovitaAI | qwen/qwen3-235b-a22b-instruct-2507 | $0.09/MTok | $0.58/MTok | 131.1K | 16.4K | |
 LLM Gateway | qwen3-235b-a22b-instruct-2507 | $0.09/MTok | $0.58/MTok | 131.1K | 8.2K | |
 Meganova | Qwen/Qwen3-235B-A22B-Instruct-2507 | $0.09/MTok | $0.60/MTok | 262K | 262K | |
 OpenRouter | qwen/qwen3-235b-a22b-2507 | $0.09/MTok | $0.10/MTok | 262.1K | 16.4K | |
 Weights & Biases | Qwen/Qwen3-235B-A22B-Instruct-2507 | $0.10/MTok | $0.10/MTok | 262.1K | 262.1K | |
 Jiekou.AI | qwen/qwen3-235b-a22b-instruct-2507 | $0.15/MTok | $0.80/MTok | 131.1K | 16.4K | |
 submodel | Qwen/Qwen3-235B-A22B-Instruct-2507 | $0.20/MTok | $0.30/MTok | 262.1K | 131.1K | |
 Nebius Token Factory | Qwen/Qwen3-235B-A22B-Instruct-2507 | $0.20/MTok | $0.60/MTok | 262.1K | 8.2K | |
 Friendli | Qwen/Qwen3-235B-A22B-Instruct-2507 | $0.20/MTok | $0.80/MTok | 262.1K | 262.1K | |
 Vercel AI Gateway | alibaba/qwen-3-235b | $0.22/MTok | $0.88/MTok | 262.1K | 16.4K | |
 302.AI | qwen3-235b-a22b-instruct-2507 | $0.29/MTok | $1.14/MTok | 128K | 65.5K | |
 Scaleway | qwen3-235b-a22b-instruct-2507 | $0.75/MTok | $2.25/MTok | 260K | 16.4K | |
 Qiniu | qwen3-235b-a22b-instruct-2507 | —/MTok | —/MTok | 262.1K | 64K |
Capabilities
Reasoning
Tool Calling
Attachments
Open Weights
Structured Output