Model comparison
DeepSeek-V2.5 (Sep 2024) vs Qwen3 235B-A22B
Qwen3 235B-A22B is the stronger model overall, scoring 43.2 to 37.0 on the Noometry Index.
Last verified . 18 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 18 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 1 category and Qwen3 235B-A22B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 235B-A22B leads 49.5 to 34.2.
- The biggest single-benchmark swing is Aider Polyglot: 17.8% for DeepSeek-V2.5 (Sep 2024) and 59.6% for Qwen3 235B-A22B.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Qwen3 235B-A22B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 37.0 | 43.2 |
| Released | 2024-09-06 | 2025-04 |
| Weights | Open | Open |
| Context window | — | 131K |
| Max output | — | 16K |
| Input $ / M tokens | — | $0.70 |
| Output $ / M tokens | — | $2.80 |
| Results tracked | 22 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 235B-A22B leads
DeepSeek-V2.5 (Sep 2024): 31.3 (#274), Qwen3 235B-A22B: 43.8 (#72)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3 235B-A22B |
|---|---|---|
| Aider Polyglot | 17.8% | 59.6% |
| LMArena Coding | 1310 | 1445 |
| SciCode | — | 42.4% |
| WeirdML | — | 41% |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, Qwen3 235B-A22B: 33.7 (#48)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3 235B-A22B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 52.1% |
| Vending-Bench 2 | — | -11.34 |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 24.5 (#145), Qwen3 235B-A22B: 15.8 (#302)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3 235B-A22B |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1433 |
| ARC-AGI-2 | — | 1.3% |
| SimpleBench | — | 31% |
| Kagi LLM Benchmark | — | 69.4% |
| ARC-AGI-1 | — | 11% |
| CritPt | — | 0% |
| Chess Puzzles | — | 12% |
| Mystery Game Puzzles | — | 9% |
| DTBench | — | 80.3% |
| LMCA | — | 29.3% |
| Epoch Capabilities Index | — | 143.87 |
| ForecastBench | — | 59.7 |
Math Qwen3 235B-A22B leads
DeepSeek-V2.5 (Sep 2024): 33.3 (#177), Qwen3 235B-A22B: 48.5 (#51)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3 235B-A22B |
|---|---|---|
| LMArena Math | 1289 | 1431 |
| OTIS Mock AIME 2024-2025 | — | 86.7% |
| Omni-MATH | — | 71.8% |
| MATH Level 5 | — | 68.9% |
| FrontierMath (Feb 2025 set) | — | 8.5% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3 235B-A22B leads
DeepSeek-V2.5 (Sep 2024): 34.2 (#190), Qwen3 235B-A22B: 49.5 (#67)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3 235B-A22B |
|---|---|---|
| LMArena Expert | 1266 | 1460 |
| GPQA Diamond | — | 80.1% |
| SimpleQA Verified | — | 40.4% |
| MMLU-Pro | — | 84.4% |
| Confabulations | — | 15.6% |
| Vectara Hallucination Rate | — | 9.3% |
| GPQA (HELM) | — | 72.7% |
Multilingual Qwen3 235B-A22B leads
DeepSeek-V2.5 (Sep 2024): 42.7 (#188), Qwen3 235B-A22B: 52.5 (#84)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3 235B-A22B |
|---|---|---|
| LMArena Non-English | 1273 | 1410 |
| LMArena Chinese | 1317 | 1481 |
| LMArena French | 1288 | 1445 |
| LMArena German | 1258 | 1432 |
| LMArena Japanese | 1228 | 1400 |
| LMArena Korean | 1209 | 1391 |
| LMArena Russian | 1289 | 1411 |
| LMArena Spanish | 1248 | 1429 |
Instruction Following Qwen3 235B-A22B leads
DeepSeek-V2.5 (Sep 2024): 67.7 (#189), Qwen3 235B-A22B: 72.8 (#132)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3 235B-A22B |
|---|---|---|
| LMArena Instruction Following | 1280 | 1408 |
| IFEval | — | 83.5% |
Long Context Qwen3 235B-A22B leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#169), Qwen3 235B-A22B: 46.1 (#25)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3 235B-A22B |
|---|---|---|
| LMArena Longer Query | 1301 | 1426 |
| Fiction.LiveBench | — | 75% |
Writing & Preference Qwen3 235B-A22B leads
DeepSeek-V2.5 (Sep 2024): 50.0 (#182), Qwen3 235B-A22B: 59.8 (#103)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3 235B-A22B |
|---|---|---|
| LMArena Text | 1294 | 1419 |
| LMArena Creative Writing | 1285 | 1384 |
| LMArena Multi-Turn | 1297 | 1432 |
| Short-Story Creative Writing | — | 83% |
| EQ-Bench Creative Writing | — | 1366 |
| WildBench | — | 86.6% |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Qwen3 235B-A22B?
Qwen3 235B-A22B is the stronger model overall, scoring 43.2 to 37.0 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Qwen3 235B-A22B better for coding?
Qwen3 235B-A22B scores higher on coding benchmarks: 43.8 versus 31.3 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Qwen3 235B-A22B share?
18 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Qwen3 235B-A22B has 49.