Qwen3.5 Flash
| Released | 2026-02-23 |
|---|---|
| Price per 1M | $0.10 in · $0.40 out |
| Overall score | 49.7 · rank 115 |
| Context window | 1,000,000 tokens |
| Provider | Alibaba |
| Weights | Proprietary |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 49.7 | 115 | 14 |
| Overall: Coding | 49.3 | 115 | 7 |
| Overall: Reasoning | 49.5 | 117 | 8 |
| Overall: Agentic work | 51.5 | 105 | 6 |
| Overall: Frontend | 41.1 | 170 | 4 |
| Coding | 52.7 | 97 | 2 |
| Agents | 50.8 | 101 | 1 |
| Reasoning | 52.3 | 95 | 3 |
| Math | 47.8 | 132 | 3 |
| Knowledge | 46.3 | 109 | 2 |
| Frontend | 33.2 | 126 | 1 |
| General | 47.9 | 129 | 1 |
| Multimodal | 53.9 | 57 | 1 |
| Value vs price | +7.2 | 50 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 49.7 ±2.0 | 14 |
| Non-reasoning | 50.6 ±8.0 | 2 |
Evidence
Coding 52.7 ±3.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| SWE-bench Verified (Vals run) | 64.4% | Vals AI | 2026-09-01 |
| LiveCodeBench (Vals run) | 83.3% | Vals AI | 2026-09-01 |
Agents 50.8 ±7.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Vending-Bench 2 | $463 | Epoch AI | 2026-11-11 |
Reasoning 52.3 ±6.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Chess Puzzles | 14.0% | Epoch AI | 2026-10-13 |
| Mystery Game Puzzles | 20.0% | Epoch AI | 2026-08-27 |
| GPQA Diamond | 82.3% | Epoch AI | 2026-08-07 |
| GPQA Diamond (Vals run) | 82.8% | Vals AI | 2026-09-01 |
Math 47.8 ±3.8
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierMath Tier 4 (2025 set) | 0.0% | Epoch AI | 2026-05-12 |
| OTIS Mock AIME 2024-2025 | 84.4% | Epoch AI | 2026-08-07 |
| FrontierMath Tiers 1-3 | 18.3% | Epoch AI | 2026-08-28 |
| FrontierMath Tiers 1-3 (2025 set) | 6.2% | Epoch AI | 2026-05-12 |
Knowledge 46.3 ±2.5
| Benchmark | Result | Source | Date |
|---|---|---|---|
| SimpleQA Verified | 20.3% | Epoch AI | 2026-08-27 |
| MMLU-Pro (Vals run) | 84.1% | Vals AI | 2026-09-01 |
Frontend 33.2 ±3.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena WebDev | 1,243 | LMArena | 2026-11-09 |
General 47.9 ±4.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Text | 1,396 | LMArena | 2026-11-09 |
Multimodal 53.9 ±2.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| MMMU-Pro (Vals run) | 81.9% | Vals AI | 2026-09-01 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| FrontierMath Tiers 1-3 | 9.5% | Epoch AI | A stronger configuration result was used |
| Mystery Game Puzzles | 16.0% | Epoch AI | A stronger configuration result was used |
Names matched to this model
qwen3-5-flash