# Qwen: Qwen3 235B A22B Instruct 2507

Provider: qwen  
Category: llm  
Model ID: `qwen/qwen3-235b-a22b-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,...

## Specs

- Context length: 262144 tokens
- Max output: unknown tokens
- Modalities: text
- Released: 2025-07-21

## Pricing

- Input: $0.07 per million tokens
- Output: $0.10 per million tokens

## Providers

- **qwen** — ctx 262144, input $0.07/M, output $0.10/M

---
Last verified: 2026-04-23T23:46:29.618Z  
Canonical URL: https://switchy.build/models/qwen3-235b-a22b-2507