Qwen3.5 Flash: benchmark scores, price and ranking
Qwen3.5 Flash from Alibaba ranks #519 of 743 LLMs on the LLMs Tiger aggregate benchmark score (46.4), based on 11 benchmark results captured up to 2026-10-13.
- Organisation
- Alibaba
- Type
- Fast / low-cost (flash) model
- Open weights
- No
- Released
- 2026-02-25 (source: epoch)
- Cheapest API price
- $0.065 in / $0.260 out per million tokens (via openrouter, observed 2026-10-13T19:47Z)
- Aggregate rank
- #519 of 743 (score 46.4)
Relative to other models, Qwen3.5 Flash is strongest on DTBench (#41 of 146) and weakest on ALE-Bench (#75 of 79).
Across the 11 benchmarks where at least ten models are scored, Qwen3.5 Flash is in the top 10% on 0 and in the top half on 4.
Its cheapest listed input price, $0.065 per million tokens, is lower than that of 89% of the 427 models with a listed price.
Qwen3.5 Flash benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| DTBench | 82.9% | #41 of 146 | Epoch AI Benchmarking Hub | 2026-10-13 |
| OTIS Mock AIME | 84.4% | #58 of 189 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Chess Puzzles | 21.0% | #43 of 122 | Epoch AI Benchmarking Hub | 2026-10-13 |
| GPQA Diamond | 82.3% | #97 of 275 | Epoch AI Benchmarking Hub | 2026-10-13 |
| LMCA | 29.1% | #55 of 108 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Mystery Games | 20.0% | #35 of 59 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Vending-Bench 2 | $463 | #15 of 19 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SimpleQA Verified | 20.3% | #56 of 69 | Epoch AI Benchmarking Hub | 2026-10-13 |
| FrontierMath T1-3 | 18.2% | #59 of 68 | Epoch AI Benchmarking Hub | 2026-10-13 |
| WebDev Arena | 1,238 | #60 of 67 | Epoch AI Benchmarking Hub | 2026-10-13 |
| ALE-Bench | 222 | #75 of 79 | Epoch AI Benchmarking Hub | 2026-10-13 |
About Qwen3.5 Flash
Qwen3.5 Flash is a smaller, lower-cost model from Alibaba. It was released on 2026-02-25, according to Epoch AI. The cheapest API listing we track charges $0.065 per million input tokens and $0.260 per million output tokens, through openrouter. We hold 11 benchmark results for it across 4 areas, and none is in the top tenth of its board.
Where Qwen3.5 Flash stands
On our aggregate score, Qwen3.5 Flash is #519 of 743 ranked models, in the lower half. Its score is 46.4; the scale is centred on 50, so it sits 3.6 points below the typical model. That beats 30% of the ranked field. Among the 74 ranked Alibaba models it is #54. The next one up is qwen3-1.7b at 46.5. The next one down is qwen-max-2025-01-25 at 46.4.
Qwen3.5 Flash, benchmark by benchmark
DTBench puts Qwen3.5 Flash at 82.9%, #41 of 146. That beats 72% of the models tested, but 16.0 points short of Claude Opus 5.5 (max) (98.9%). The results either side are Qwen3.7 Plus (84.0%) and Gemma 4 31B IT (82.7%). It covers dynamic tool-use benchmark. The board median is 68.8%, so Qwen3.5 Flash is 14.1 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On OTIS Mock AIME, Qwen3.5 Flash reaches 84.4%. It ranks #58 of 189, ahead of 69% of the field, and 15.6 points behind Claude Fable 5.1 (max) at 100.0%. The results either side are Qwen 3.6 Flash (2026-04-16) (84.4%) and grok-4-0709 (84.0%). It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so Qwen3.5 Flash is 26.7 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Qwen3.5 Flash manages 21.0% on Chess Puzzles, good for #43 of 122. GPT-6 Astra (max) leads at 72.0%, 51.0 points further up, and 65% of models sit below. The results either side are Kimi K2.7 Code (21.0%) and DeepSeek v4 Pro (max) (20.0%). It covers chess tactics puzzles. The board median is 12.5%, so Qwen3.5 Flash is 8.5 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
GPQA Diamond puts Qwen3.5 Flash at 82.3%, #97 of 275. That beats 65% of the models tested, but 13.4 points short of GPT-6 Astra (max) (95.8%). The results either side are Claude Sonnet 4.5 (59k thinking) (82.3%) and Qwen3.7 Flash (82.3%). It covers expert-level science questions. The board median is 71.7%, so Qwen3.5 Flash is 10.6 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Qwen3.5 Flash lands below the middle on LMCA: 29.1%, #55 of 108, with 50% of the field scoring higher. Claude Opus 5.5 (max) leads at 68.2%; the closest result above is Qwen3-235B-A22B-Thinking-2507 at 29.3%. It covers language-model capability assessment. The board median is 29.2%, so Qwen3.5 Flash is 0.1 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Qwen3.5 Flash lands below the middle on Mystery Games: 20.0%, #35 of 59, with 58% of the field scoring higher. GPT-6 Astra (max) leads at 84.0%; the closest result above is nemotron-3-ultra at 20.0%. It covers deductive reasoning puzzles. The board median is 23.0%, so Qwen3.5 Flash is 3.0 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
SimpleQA Verified is a weak spot. Qwen3.5 Flash scores 20.3%, #56 of 69, behind 80% of the models measured. The leader, gemini-4-argon-high, reaches 76.0%. It covers short factual questions. The board median is 41.0%, so Qwen3.5 Flash is 20.7 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
FrontierMath T1-3 is a weak spot. Qwen3.5 Flash scores 18.2%, #59 of 68, behind 85% of the models measured. The leader, GPT-6 Astra (max), reaches 93.7%. It covers research-level maths, tiers 1-3. The board median is 55.6%, so Qwen3.5 Flash is 37.4 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
WebDev Arena is a weak spot. Qwen3.5 Flash scores 1,238, #60 of 67, behind 88% of the models measured. The leader, Claude Opus 5.5 (max), reaches 1,820. It covers building web apps, crowd-voted (Elo). The board median is 1,433, so Qwen3.5 Flash is 195 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
ALE-Bench is a weak spot. Qwen3.5 Flash scores 222, #75 of 79, behind 94% of the models measured. The leader, GPT-6 Astra (max), reaches 2,951. It covers AtCoder heuristic-optimisation contests (performance rating). The board median is 764, so Qwen3.5 Flash is 542 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
The full table above also lists the other 1 result, including boards with fewer than ten models, where a rank says little.
Strengths and weaknesses
Software engineering: weak, ahead of 8% of the field on average across 2 benchmarks. Its best board here is WebDev Arena (#60 of 67) and its weakest is ALE-Bench (#75 of 79). Reasoning, maths and science: uneven, ahead of 49% of the field on average across 3 benchmarks. Its best board here is OTIS Mock AIME (#58 of 189) and its weakest is FrontierMath T1-3 (#59 of 68). General knowledge and chat: uneven, ahead of 39% of the field on average across 5 benchmarks. Its best board here is Chess Puzzles (#43 of 122) and its weakest is SimpleQA Verified (#56 of 69).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so Qwen3.5 Flash costs less than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen2.5-Coder-3B-Instruct ($0.010, 46.5). Similar scores cost more on DeepSeek-R1-Distill-Qwen-14B ($0.070) and phi-4 ($0.070).
Release history
The Alibaba model released before it was Qwen3.5 9B on 2026-02-24, which scores 44.9. The next one was Qwen 3.6 Plus (2026-04-02) on 2026-04-02, at 50.1.
Reading these numbers
The 11 scores on this page come from 1 source: Epoch AI Benchmarking Hub (11). No official leaderboard supplied any of them, so treat them as third-party measurements.
Models ranked nearby
- #516 deepseek-chat (46.5)
- #517 Qwen2.5-7B (46.5)
- #518 qwen3-1.7b (46.5)
- #520 qwen-max-2025-01-25 (46.4)
- #521 Command R (08-24) (46.4)
- #522 Command R+ (08-24) (46.4)
More from Alibaba
- Qwen 3.6 Max (Preview) (#25, 54.1)
- Qwen3.8 Max (xhigh) (#37, 53.2)
- Qwen3.7 Max (#71, 52.3)
- ollama/qwen2.5-coder:32b (#93, 51.9)
- Qwen3 235B A22B diff, no think, Alibaba API (#122, 51.5)
- Qwen2.5-Coder-14B-Instruct (#135, 51.3)
Compare it on the live leaderboard → · How the aggregate score works