Model comparison
DeepSeek-V2.5 (Sep 2024) vs Qwen3.8 27B
Qwen3.8 27B is the stronger model overall, scoring 46.0 to 37.6 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 0 categories and Qwen3.8 27B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in coding, where Qwen3.8 27B leads 50.5 to 31.7.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 37.6 | 46.0 |
| Released | 2024-09-06 | 2026-08-14 |
| Weights | Open | Open |
| Context window | — | 262K |
| Max output | — | 33K |
| Input $ / M tokens | — | $0.99 |
| Output $ / M tokens | — | $1.49 |
| Results tracked | 22 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 27B leads
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), Qwen3.8 27B: 50.5 (#44)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B |
|---|---|---|
| LMArena Coding | 1309 | 1482 |
| Aider Polyglot | 17.8% | — |
| LMArena WebDev | — | 1593 |
| SciCode | — | 46.6% |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, Qwen3.8 27B: 32.9 (#57)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B |
|---|---|---|
| APEX-Agents | — | 47.5% |
Reasoning Qwen3.8 27B leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), Qwen3.8 27B: 41.0 (#54)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1460 |
| ARC-AGI-2 | — | 42.4% |
| NYT Connections (extended) | — | 54.5% |
| ARC-AGI-1 | — | 87.5% |
| CritPt | — | 5.4% |
| DTBench | — | 88% |
| LMCA | — | 41.4% |
| Surface Evolver Bench | — | 45% |
| Epoch Capabilities Index | — | 149.38 |
Math Qwen3.8 27B leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), Qwen3.8 27B: 37.1 (#161)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B |
|---|---|---|
| LMArena Math | 1288 | 1456 |
| ProofBench | — | 16% |
Knowledge Qwen3.8 27B leads
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), Qwen3.8 27B: 41.6 (#109)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B |
|---|---|---|
| LMArena Expert | 1266 | 1482 |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, Qwen3.8 27B: 41.3 (#37)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B |
|---|---|---|
| LMArena Vision | — | 1271 |
Multilingual Qwen3.8 27B leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), Qwen3.8 27B: 53.7 (#60)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B |
|---|---|---|
| LMArena Non-English | 1273 | 1430 |
| LMArena Chinese | 1318 | 1504 |
| LMArena French | 1289 | 1465 |
| LMArena German | 1258 | 1438 |
| LMArena Japanese | 1228 | 1384 |
| LMArena Korean | 1209 | 1393 |
| LMArena Russian | 1289 | 1415 |
| LMArena Spanish | 1248 | 1448 |
Instruction Following Qwen3.8 27B leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), Qwen3.8 27B: 75.8 (#53)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B |
|---|---|---|
| LMArena Instruction Following | 1280 | 1439 |
Long Context Qwen3.8 27B leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), Qwen3.8 27B: 44.3 (#70)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B |
|---|---|---|
| LMArena Longer Query | 1301 | 1450 |
Writing & Preference Qwen3.8 27B leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), Qwen3.8 27B: 65.8 (#43)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 27B |
|---|---|---|
| LMArena Text | 1294 | 1441 |
| LMArena Creative Writing | 1285 | 1384 |
| LMArena Multi-Turn | 1297 | 1441 |
| EQ-Bench Creative Writing | — | 1671 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Qwen3.8 27B?
Qwen3.8 27B is the stronger model overall, scoring 46.0 to 37.6 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Qwen3.8 27B better for coding?
Qwen3.8 27B scores higher on coding benchmarks: 50.5 versus 31.7 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Qwen3.8 27B share?
17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Qwen3.8 27B has 31.