Qwen3 235B-A22B Instruct 2507
| Released | 2025-07-21 |
|---|---|
| Price per 1M | $0.14 in · $0.80 out |
| Overall score | 45.5 · rank 154 |
| Context window | 262,144 tokens |
| Provider | Alibaba |
| Weights | Open weights (Apache 2.0) |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 45.5 | 154 | 6 |
| Overall: Coding | 44.6 | 159 | 4 |
| Overall: Reasoning | 44.4 | 164 | 2 |
| Overall: Agentic work | 44.4 | 163 | 4 |
| Overall: Frontend | 44.8 | 139 | 1 |
| Coding | 44.8 | 157 | 1 |
| Agents | 44.3 | 139 | 1 |
| Reasoning | 44.4 | 159 | 2 |
| General | 51.0 | 93 | 2 |
| Value vs price | +1.2 | 111 |
Evidence
Coding 44.8 ±1.5
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Aider Polyglot | 59.6% | Epoch AI | 2026-11-11 |
Agents 44.3 ±3.6
Reasoning 44.4 ±2.0
General 51.0 ±3.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Text | 1,422 | LMArena | 2026-11-09 |
| EQ-Bench Longform Creative Writing | 57.4% | EQ-Bench | 2026-11-11 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| BFCL v4 | 48.0% | BFCL | A stronger configuration result was used |
Names matched to this model
parasail-qwen3-235b-a22b-instruct-2507, qwen3-235b-a22b-instruct-2507