Model comparison
DeepSeek-V2.5 (Sep 2024) vs Qwen3-30B-A3B
Qwen3-30B-A3B is the stronger model overall, scoring 38.5 to 37.0 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 17 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 2 categories and Qwen3-30B-A3B in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where DeepSeek-V2.5 (Sep 2024) leads 39.5 to 31.0.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Qwen3-30B-A3B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 37.0 | 38.5 |
| Released | 2024-09-06 | 2025-04-28 |
| Weights | Open | Open |
| Context window | — | 41K |
| Max output | — | 16K |
| Input $ / M tokens | — | $0.12 |
| Output $ / M tokens | — | $0.50 |
| Results tracked | 22 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3-30B-A3B leads
DeepSeek-V2.5 (Sep 2024): 31.3 (#274), Qwen3-30B-A3B: 37.0 (#188)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3-30B-A3B |
|---|---|---|
| LMArena Coding | 1310 | 1416 |
| Aider Polyglot | 17.8% | — |
| SciCode | — | 33.3% |
| WeirdML | — | 29.8% |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, Qwen3-30B-A3B: 29.5 (#75)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3-30B-A3B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 41.4% |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 24.5 (#145), Qwen3-30B-A3B: 21.6 (#193)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3-30B-A3B |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1398 |
| Kagi LLM Benchmark | — | 54.9% |
| CritPt | — | 0.3% |
| Chess Puzzles | — | 8% |
| DTBench | — | 69.3% |
| LMCA | — | 22.4% |
| Epoch Capabilities Index | — | 139.63 |
Math Qwen3-30B-A3B leads
DeepSeek-V2.5 (Sep 2024): 33.3 (#177), Qwen3-30B-A3B: 36.0 (#133)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3-30B-A3B |
|---|---|---|
| LMArena Math | 1289 | 1394 |
| MathArena Final-Answer Competitions | — | 47.8% |
| OTIS Mock AIME 2024-2025 | — | 70.3% |
Knowledge Qwen3-30B-A3B leads
DeepSeek-V2.5 (Sep 2024): 34.2 (#190), Qwen3-30B-A3B: 41.7 (#96)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3-30B-A3B |
|---|---|---|
| LMArena Expert | 1266 | 1395 |
| GPQA Diamond | — | 70.1% |
| Confabulations | — | 12.3% |
Multilingual Qwen3-30B-A3B leads
DeepSeek-V2.5 (Sep 2024): 42.7 (#188), Qwen3-30B-A3B: 49.8 (#126)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3-30B-A3B |
|---|---|---|
| LMArena Non-English | 1273 | 1372 |
| LMArena Chinese | 1317 | 1434 |
| LMArena French | 1288 | 1419 |
| LMArena German | 1258 | 1379 |
| LMArena Japanese | 1228 | 1338 |
| LMArena Korean | 1209 | 1333 |
| LMArena Russian | 1289 | 1369 |
| LMArena Spanish | 1248 | 1403 |
Instruction Following Qwen3-30B-A3B leads
DeepSeek-V2.5 (Sep 2024): 67.7 (#189), Qwen3-30B-A3B: 72.2 (#137)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3-30B-A3B |
|---|---|---|
| LMArena Instruction Following | 1280 | 1364 |
Long Context DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#169), Qwen3-30B-A3B: 31.0 (#278)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3-30B-A3B |
|---|---|---|
| LMArena Longer Query | 1301 | 1379 |
| Fiction.LiveBench | — | 40.6% |
Writing & Preference Qwen3-30B-A3B leads
DeepSeek-V2.5 (Sep 2024): 50.0 (#182), Qwen3-30B-A3B: 55.9 (#138)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3-30B-A3B |
|---|---|---|
| LMArena Text | 1294 | 1384 |
| LMArena Creative Writing | 1285 | 1318 |
| LMArena Multi-Turn | 1297 | 1378 |
| Short-Story Creative Writing | — | 75.3% |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Qwen3-30B-A3B?
Qwen3-30B-A3B is the stronger model overall, scoring 38.5 to 37.0 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Qwen3-30B-A3B better for coding?
Qwen3-30B-A3B scores higher on coding benchmarks: 37.0 versus 31.3 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Qwen3-30B-A3B share?
17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Qwen3-30B-A3B has 32.