Model comparison
DeepSeek V4 Flash vs Qwen3.5-Flash
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 42.5 on the Noometry Index.
Last verified . 28 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 28 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 8 categories and Qwen3.5-Flash in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4 Flash leads 60.3 to 37.4.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 57.5% for DeepSeek V4 Flash and 18.2% for Qwen3.5-Flash.
- Qwen3.5-Flash is cheaper at $0.10 / $0.40 per million input/output tokens, against $0.15 / $0.60 for DeepSeek V4 Flash.
- DeepSeek V4 Flash has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4 Flash | Qwen3.5-Flash | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 53.6 | 42.5 |
| Released | 2026-04-24 | 2026-02-23 |
| Weights | Open | Proprietary |
| Context window | 1M | 1M |
| Max output | 393K | 66K |
| Input $ / M tokens | $0.15 | $0.10 |
| Output $ / M tokens | $0.60 | $0.40 |
| Results tracked | 41 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Flash leads
DeepSeek V4 Flash: 47.9 (#59), Qwen3.5-Flash: 34.2 (#242)
| Benchmark | DeepSeek V4 Flash | Qwen3.5-Flash |
|---|---|---|
| LMArena WebDev | 1582 | 1244 |
| LMArena Coding | 1457 | 1412 |
| ALE-Bench | 1,306 | 221.8 |
| FrontierCode | 18.8% | — |
| SciCode | 49.9% | — |
| WeirdML | 63% | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Flash: —, Qwen3.5-Flash: —
| Benchmark | DeepSeek V4 Flash | Qwen3.5-Flash |
|---|---|---|
| Vending-Bench 2 | — | 462.69 |
Reasoning DeepSeek V4 Flash leads
DeepSeek V4 Flash: 53.7 (#30), Qwen3.5-Flash: 33.7 (#72)
| Benchmark | DeepSeek V4 Flash | Qwen3.5-Flash |
|---|---|---|
| Chess Puzzles | 33% | 21% |
| LMArena Hard Prompts | 1444 | 1403 |
| Mystery Game Puzzles | 34% | 20% |
| DTBench | 90.9% | 82.9% |
| LMCA | 41.7% | 29.1% |
| Epoch Capabilities Index | 154.49 | 143.98 |
| ARC-AGI-2 | 61.4% | — |
| SimpleBench | 61.1% | — |
| Kagi LLM Benchmark | 52.2% | — |
| NYT Connections (extended) | 89.6% | — |
| ARC-AGI-1 | 89% | — |
| CritPt | 16.6% | — |
Math DeepSeek V4 Flash leads
DeepSeek V4 Flash: 60.3 (#37), Qwen3.5-Flash: 37.4 (#158)
| Benchmark | DeepSeek V4 Flash | Qwen3.5-Flash |
|---|---|---|
| FrontierMath (Tiers 1-3) | 57.5% | 18.2% |
| OTIS Mock AIME 2024-2025 | 94.4% | 84.4% |
| LMArena Math | 1427 | 1407 |
| FrontierMath Tier 4 | 24.4% | — |
| MathArena Final-Answer Competitions | 76.5% | — |
| ProofBench | 56% | — |
| FrontierMath (Feb 2025 set) | — | 6.2% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge DeepSeek V4 Flash leads
DeepSeek V4 Flash: 55.4 (#48), Qwen3.5-Flash: 43.2 (#93)
| Benchmark | DeepSeek V4 Flash | Qwen3.5-Flash |
|---|---|---|
| GPQA Diamond | 91% | 82.3% |
| SimpleQA Verified | 33.6% | 20.3% |
| LMArena Expert | 1441 | 1407 |
| Vectara Hallucination Rate | — | 10.5% |
Multilingual DeepSeek V4 Flash leads
DeepSeek V4 Flash: 53.0 (#72), Qwen3.5-Flash: 50.5 (#121)
| Benchmark | DeepSeek V4 Flash | Qwen3.5-Flash |
|---|---|---|
| LMArena Non-English | 1420 | 1385 |
| LMArena Chinese | 1468 | 1446 |
| LMArena French | 1439 | 1412 |
| LMArena German | 1418 | 1390 |
| LMArena Japanese | 1406 | 1368 |
| LMArena Korean | 1384 | 1344 |
| LMArena Russian | 1428 | 1379 |
| LMArena Spanish | 1436 | 1400 |
Instruction Following DeepSeek V4 Flash leads
DeepSeek V4 Flash: 74.9 (#81), Qwen3.5-Flash: 72.6 (#139)
| Benchmark | DeepSeek V4 Flash | Qwen3.5-Flash |
|---|---|---|
| LMArena Instruction Following | 1421 | 1374 |
Long Context DeepSeek V4 Flash leads
DeepSeek V4 Flash: 43.8 (#85), Qwen3.5-Flash: 42.4 (#124)
| Benchmark | DeepSeek V4 Flash | Qwen3.5-Flash |
|---|---|---|
| LMArena Longer Query | 1434 | 1392 |
Writing & Preference DeepSeek V4 Flash leads
DeepSeek V4 Flash: 63.8 (#61), Qwen3.5-Flash: 57.9 (#122)
| Benchmark | DeepSeek V4 Flash | Qwen3.5-Flash |
|---|---|---|
| LMArena Text | 1432 | 1397 |
| LMArena Creative Writing | 1403 | 1343 |
| LMArena Multi-Turn | 1449 | 1393 |
| EQ-Bench Creative Writing | 1559 | — |
Frequently asked questions
Is DeepSeek V4 Flash better than Qwen3.5-Flash?
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 42.5 on the Noometry Index.
Which is cheaper, DeepSeek V4 Flash or Qwen3.5-Flash?
Qwen3.5-Flash is cheaper. It lists at $0.10 per million input tokens and $0.40 per million output tokens; DeepSeek V4 Flash lists at $0.15 and $0.60.
Is DeepSeek V4 Flash or Qwen3.5-Flash better for coding?
DeepSeek V4 Flash scores higher on coding benchmarks: 47.9 versus 34.2 in the Noometry coding category.
Which has the bigger context window?
Both accept 1M tokens.
How many benchmarks do DeepSeek V4 Flash and Qwen3.5-Flash share?
28 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and Qwen3.5-Flash has 32.