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.
function_calling| Date | Input | Output | Cache read | Change |
|---|---|---|---|---|
| Mar 21, 2026 | $0.071 | $0.1 | — | — |
| Jun 10, 2026d68a4f1 | $0.09 | $0.1 | — | ↑27% |
| Jul 17, 202620c73ac | $0.09 | $0.55 | — | ↑0.0% |
| Aug 4, 202680c184a | $0.15 | $0.598 | — | ↑67% |
| Aug 13, 20263cddff5 | $0.09 | $0.55 | — | ↓40% |
| Aug 31, 2026e3d7ded | $0.088 | $0.35 | $0.018 | ↓2% |
| Sep 1, 2026639b255 | $0.09 | $0.55 | $0.018 | ↑2% |
| Sep 2, 202677b8547 | $0.088 | $0.35 | $0.018 | ↓2% |
| Sep 6, 20265cd8ca9 | $0.09 | $0.55 | $0.018 | ↑2% |
| Sep 9, 202679684d6 | $0.22 | $0.88 | $0.018 | ↑144% |
| Sep 12, 2026265df08 | $0.088 | $0.35 | $0.018 | ↓60% |
| Oct 6, 202600710a0 | $0.09 | $0.55 | $0.018 | ↑2% |
Input is ↑27% since Mar 21, 2026, from $0.071 to $0.09 per 1M tokens.
| Model | Context | Max out | Pricing |
|---|---|---|---|
| Qwen: Qwen3 4B (free) | 41K | — | |
| Qwen: Qwen3 Coder 480B A35B (free) | 1.0M | 262K | |
| Qwen: Qwen3 Next 80B A3B Instruct (free) | 262K | — | |
| Qwen: Qwen3 Coder 30B A3B Instruct | 262K | 236K | $0.07$0.28 |
| Qwen: Qwen3 32B | 131K | 16K | $0.08$0.28 |
| Qwen: Qwen3 235B A22B Instruct 2507 | 262K | 16K | $0.09$0.55 |
| Qwen: Qwen3 14B | 41K | 37K | $0.1$0.22 |
| Qwen: Qwen3 30B A3B Instruct 2507 | 262K | 236K | $0.1$0.3 |
| Qwen: Qwen3 Next 80B A3B Instruct | 262K | 236K | $0.1$1.1 |
| Qwen: Qwen3 VL 32B Instruct | 131K | 33K | $0.1$0.42 |
| Qwen: Qwen3 8B | 131K | 8K | $0.12$0.46 |
| Qwen: Qwen3 30B A3B | 41K | 16K | $0.12$0.5 |
| Qwen: Qwen3 Coder Next | 262K | 236K | $0.12$0.8 |
| Qwen: Qwen3 Next 80B A3B Thinking | 262K | 236K | $0.15$1.2 |
| Qwen: Qwen3 VL 30B A3B Instruct | 262K | 16K | $0.15$0.6 |
| Qwen: Qwen3 VL 8B Thinking | 131K | 33K | $0.18$2.1 |
| Qwen: Qwen3 Coder Flash | 1M | 66K | $0.2$0.98 |
| Qwen: Qwen3 30B A3B Thinking 2507 | 82K | 33K | $0.2$2.4 |
| Qwen: Qwen3 VL 235B A22B Instruct | 262K | 33K | $0.21$1.9 |
| Qwen: Qwen3 VL 8B Instruct | 262K | 236K | $0.25$0.75 |
| Qwen: Qwen3 VL 30B A3B Thinking | 262K | 236K | $0.29$1 |
| Qwen: Qwen3 Coder 480B A35B | 262K | 66K | $0.3$1 |
| Qwen: Qwen3 VL 235B A22B Thinking | 131K | 33K | $0.4$4 |
| Qwen: Qwen3 235B A22B Thinking 2507 | 128K | 16K | $0.45$3.5 |
| Qwen: Qwen3 Coder Plus | 1M | 66K | $0.65$3.25 |
| Qwen: Qwen3 Max Thinking | 262K | 66K | $0.78$3.9 |
| Qwen: Qwen3 Max | 262K | 66K | $0.78$3.9 |
/v1/models/openrouter/qwen/qwen3-235b-a22b-2507Qwen3-235B-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.