Model comparison
DeepSeek-V3.1-Terminus vs Qwen2.5-Coder-32B
DeepSeek-V3.1-Terminus is the stronger model overall, scoring 42.6 to 32.9 on the Noometry Index.
Last verified . 10 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 10 benchmarks with published results for both. DeepSeek-V3.1-Terminus scores higher in 7 categories and Qwen2.5-Coder-32B in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V3.1-Terminus leads 61.3 to 41.8.
- DeepSeek-V3.1-Terminus is cheaper at $0.30 / $1 per million input/output tokens, against $0.66 / $1 for Qwen2.5-Coder-32B.
- DeepSeek-V3.1-Terminus accepts more context: 164K tokens versus 33K.
Side by side
| DeepSeek-V3.1-Terminus | Qwen2.5-Coder-32B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 42.6 | 32.9 |
| Released | 2025-09-22 | 2024-09-18 |
| Weights | Open | Open |
| Context window | 164K | 33K |
| Max output | 66K | 29K |
| Input $ / M tokens | $0.30 | $0.66 |
| Output $ / M tokens | $1 | $1 |
| Results tracked | 16 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 41.5 (#114), Qwen2.5-Coder-32B: 22.4 (#328)
| Benchmark | DeepSeek-V3.1-Terminus | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Coding | 1426 | 1276 |
| SWE-bench Verified (bash only) | — | 9% |
| Aider Polyglot | — | 16.4% |
| SciCode | 40.6% | — |
| BigCodeBench Instruct | — | 49% |
| LiveBench Coding | — | 56.9% |
| BigCodeBench Complete | — | 58% |
| ALE-Bench | 745.17 | — |
| HumanEval+ | — | 87.2% |
| MBPP+ | — | 77% |
Reasoning DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 25.7 (#126), Qwen2.5-Coder-32B: 20.4 (#220)
| Benchmark | DeepSeek-V3.1-Terminus | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Hard Prompts | 1427 | 1251 |
| Kagi LLM Benchmark | 57.4% | — |
| CritPt | 1.7% | — |
| LiveBench Reasoning | — | 42.1% |
| DTBench | 81.3% | — |
| LiveBench Data Analysis | — | 49.9% |
| LMCA | 28.6% | — |
| Epoch Capabilities Index | — | 119.41 |
| HellaSwag | — | 83% |
| LiveBench | — | 46.2% |
| WinoGrande | — | 80.8% |
Math DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 35.7 (#138), Qwen2.5-Coder-32B: 30.8 (#203)
| Benchmark | DeepSeek-V3.1-Terminus | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Math | 1402 | 1251 |
| LiveBench Math | — | 46.6% |
| GSM8K | — | 93% |
Knowledge Not comparable
DeepSeek-V3.1-Terminus: —, Qwen2.5-Coder-32B: 32.9 (#200)
| Benchmark | DeepSeek-V3.1-Terminus | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Expert | — | 1221 |
| ARC (AI2) Challenge | — | 70.5% |
| MMLU | — | 79.1% |
Multilingual DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 52.3 (#87), Qwen2.5-Coder-32B: 38.0 (#230)
| Benchmark | DeepSeek-V3.1-Terminus | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Non-English | 1407 | 1205 |
| LMArena Russian | 1434 | 1228 |
| LMArena Chinese | — | 1222 |
Instruction Following DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 74.4 (#96), Qwen2.5-Coder-32B: 61.6 (#240)
| Benchmark | DeepSeek-V3.1-Terminus | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Instruction Following | 1406 | 1223 |
| LiveBench Instruction Following | — | 58.7% |
Long Context DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 43.5 (#88), Qwen2.5-Coder-32B: 37.9 (#204)
| Benchmark | DeepSeek-V3.1-Terminus | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Longer Query | 1426 | 1251 |
Writing & Preference DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 61.3 (#87), Qwen2.5-Coder-32B: 41.8 (#235)
| Benchmark | DeepSeek-V3.1-Terminus | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Text | 1420 | 1230 |
| LMArena Creative Writing | 1404 | 1174 |
| LMArena Multi-Turn | 1411 | 1222 |
| LiveBench Language | — | 23.3% |
Frequently asked questions
Is DeepSeek-V3.1-Terminus better than Qwen2.5-Coder-32B?
DeepSeek-V3.1-Terminus is the stronger model overall, scoring 42.6 to 32.9 on the Noometry Index.
Which is cheaper, DeepSeek-V3.1-Terminus or Qwen2.5-Coder-32B?
DeepSeek-V3.1-Terminus is cheaper. It lists at $0.30 per million input tokens and $1 per million output tokens; Qwen2.5-Coder-32B lists at $0.66 and $1.
Is DeepSeek-V3.1-Terminus or Qwen2.5-Coder-32B better for coding?
DeepSeek-V3.1-Terminus scores higher on coding benchmarks: 41.5 versus 22.4 in the Noometry coding category.
Which has the bigger context window?
DeepSeek-V3.1-Terminus does, with 164K tokens against 33K.
How many benchmarks do DeepSeek-V3.1-Terminus and Qwen2.5-Coder-32B share?
10 benchmarks have published results for both models. DeepSeek-V3.1-Terminus has 16 scored results on Noometry and Qwen2.5-Coder-32B has 31.