Model comparison
DeepSeek V4 Flash vs Qwen2.5 7B Instruct
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 29.0 on the Noometry Index.
Last verified . 6 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 6 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 6 categories and Qwen2.5 7B Instruct in 0 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4 Flash leads 60.3 to 12.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 94.4% for DeepSeek V4 Flash and 2.5% for Qwen2.5 7B Instruct.
- DeepSeek V4 Flash is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.17 / $0.70 for Qwen2.5 7B Instruct.
- DeepSeek V4 Flash accepts more context: 1M tokens versus 131K.
Side by side
| DeepSeek V4 Flash | Qwen2.5 7B Instruct | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 53.6 | 29.0 |
| Released | 2026-04-24 | 2024-09 |
| Weights | Open | Open |
| Context window | 1M | 131K |
| Max output | 393K | 8K |
| Input $ / M tokens | $0.15 | $0.17 |
| Output $ / M tokens | $0.60 | $0.70 |
| Results tracked | 41 | 15 |
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), Qwen2.5 7B Instruct: 36.5 (#208)
| Benchmark | DeepSeek V4 Flash | Qwen2.5 7B Instruct |
|---|---|---|
| FrontierCode | 18.8% | — |
| LMArena WebDev | 1582 | — |
| SciCode | 49.9% | — |
| WeirdML | 63% | — |
| BigCodeBench Instruct | — | 37.6% |
| LMArena Coding | 1457 | — |
| BigCodeBench Complete | — | 46.1% |
| ALE-Bench | 1,306 | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Flash: —, Qwen2.5 7B Instruct: 23.8 (#124)
| Benchmark | DeepSeek V4 Flash | Qwen2.5 7B Instruct |
|---|---|---|
| BALROG | — | 7.8% |
Reasoning DeepSeek V4 Flash leads
DeepSeek V4 Flash: 53.7 (#30), Qwen2.5 7B Instruct: 14.8 (#322)
| Benchmark | DeepSeek V4 Flash | Qwen2.5 7B Instruct |
|---|---|---|
| Chess Puzzles | 33% | 0% |
| DTBench | 90.9% | 47.7% |
| LMCA | 41.7% | 6.4% |
| Epoch Capabilities Index | 154.49 | 118.51 |
| 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% | — |
| LMArena Hard Prompts | 1444 | — |
| Mystery Game Puzzles | 34% | — |
Math DeepSeek V4 Flash leads
DeepSeek V4 Flash: 60.3 (#37), Qwen2.5 7B Instruct: 12.6 (#306)
| Benchmark | DeepSeek V4 Flash | Qwen2.5 7B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 94.4% | 2.5% |
| FrontierMath (Tiers 1-3) | 57.5% | — |
| FrontierMath Tier 4 | 24.4% | — |
| MathArena Final-Answer Competitions | 76.5% | — |
| ProofBench | 56% | — |
| Omni-MATH | — | 29.4% |
| LMArena Math | 1427 | — |
Knowledge DeepSeek V4 Flash leads
DeepSeek V4 Flash: 55.4 (#48), Qwen2.5 7B Instruct: 17.0 (#286)
| Benchmark | DeepSeek V4 Flash | Qwen2.5 7B Instruct |
|---|---|---|
| GPQA Diamond | 91% | 35.5% |
| SimpleQA Verified | 33.6% | — |
| MMLU-Pro | — | 53.9% |
| GPQA (HELM) | — | 34.1% |
| LMArena Expert | 1441 | — |
| MMLU | — | 72.9% |
Multilingual Not comparable
DeepSeek V4 Flash: 53.0 (#72), Qwen2.5 7B Instruct: —
| Benchmark | DeepSeek V4 Flash | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Non-English | 1420 | — |
| LMArena Chinese | 1468 | — |
| LMArena French | 1439 | — |
| LMArena German | 1418 | — |
| LMArena Japanese | 1406 | — |
| LMArena Korean | 1384 | — |
| LMArena Russian | 1428 | — |
| LMArena Spanish | 1436 | — |
Instruction Following DeepSeek V4 Flash leads
DeepSeek V4 Flash: 74.9 (#81), Qwen2.5 7B Instruct: 63.2 (#231)
| Benchmark | DeepSeek V4 Flash | Qwen2.5 7B Instruct |
|---|---|---|
| IFEval | — | 74.1% |
| LMArena Instruction Following | 1421 | — |
Long Context Not comparable
DeepSeek V4 Flash: 43.8 (#85), Qwen2.5 7B Instruct: —
| Benchmark | DeepSeek V4 Flash | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Longer Query | 1434 | — |
Writing & Preference DeepSeek V4 Flash leads
DeepSeek V4 Flash: 63.8 (#61), Qwen2.5 7B Instruct: 48.8 (#195)
| Benchmark | DeepSeek V4 Flash | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Text | 1432 | — |
| LMArena Creative Writing | 1403 | — |
| EQ-Bench Creative Writing | 1559 | — |
| WildBench | — | 73.1% |
| LMArena Multi-Turn | 1449 | — |
Frequently asked questions
Is DeepSeek V4 Flash better than Qwen2.5 7B Instruct?
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 29.0 on the Noometry Index.
Which is cheaper, DeepSeek V4 Flash or Qwen2.5 7B Instruct?
DeepSeek V4 Flash is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Qwen2.5 7B Instruct lists at $0.17 and $0.70.
Is DeepSeek V4 Flash or Qwen2.5 7B Instruct better for coding?
DeepSeek V4 Flash scores higher on coding benchmarks: 47.9 versus 36.5 in the Noometry coding category.
Which has the bigger context window?
DeepSeek V4 Flash does, with 1M tokens against 131K.
How many benchmarks do DeepSeek V4 Flash and Qwen2.5 7B Instruct share?
6 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and Qwen2.5 7B Instruct has 15.