Model comparison
DeepSeek-V3.1-Terminus vs Mistral Large
DeepSeek-V3.1-Terminus is the stronger model overall, scoring 43.1 to 31.9 on the Noometry Index.
Last verified . 15 shared benchmarks.
DeepSeek
Mistral AISummary
- They share 15 benchmarks with published results for both. DeepSeek-V3.1-Terminus scores higher in 7 categories and Mistral Large in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek-V3.1-Terminus leads 38.5 to 18.2.
- The biggest single-benchmark swing is DTBench: 81.3% for DeepSeek-V3.1-Terminus and 65.1% for Mistral Large.
- DeepSeek-V3.1-Terminus is cheaper at $0.27 / $1 per million input/output tokens, against $2 / $6 for Mistral Large.
- DeepSeek-V3.1-Terminus accepts more context: 164K tokens versus 131K.
Side by side
| DeepSeek-V3.1-Terminus | Mistral Large | |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 43.1 | 31.9 |
| Released | 2025-09-22 | 2024-02-26 |
| Weights | Open | Open |
| Context window | 164K | 131K |
| Max output | 147K | 16K |
| Input $ / M tokens | $0.27 | $2 |
| Output $ / M tokens | $1 | $6 |
| Results tracked | 16 | 51 |
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: 42.0 (#113), Mistral Large: 34.3 (#240)
| Benchmark | DeepSeek-V3.1-Terminus | Mistral Large |
|---|---|---|
| SciCode | 40.6% | 36.2% |
| LMArena Coding | 1426 | 1277 |
| ALE-Bench | 745.17 | 264.7 |
| BigCodeBench Instruct | — | 30% |
| LiveBench Coding | — | 47.1% |
| BigCodeBench Complete | — | 38.3% |
| HumanEval+ | — | 62.2% |
| MBPP+ | — | 59.5% |
Agentic & Tool Use Not comparable
DeepSeek-V3.1-Terminus: —, Mistral Large: 28.6 (#89)
| Benchmark | DeepSeek-V3.1-Terminus | Mistral Large |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 38.4% |
Reasoning DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 26.4 (#133), Mistral Large: 15.8 (#310)
| Benchmark | DeepSeek-V3.1-Terminus | Mistral Large |
|---|---|---|
| CritPt | 1.7% | 0% |
| LMArena Hard Prompts | 1426 | 1257 |
| DTBench | 81.3% | 65.1% |
| LMCA | 28.6% | 16.7% |
| SimpleBench | — | 22.5% |
| Kagi LLM Benchmark | 57.4% | — |
| LiveBench Reasoning | — | 43.5% |
| LiveBench Data Analysis | — | 50.1% |
| Epoch Capabilities Index | — | 128.52 |
| ForecastBench | — | 57.1 |
| LiveBench | — | 48.4% |
Math DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 38.5 (#137), Mistral Large: 18.2 (#291)
| Benchmark | DeepSeek-V3.1-Terminus | Mistral Large |
|---|---|---|
| LMArena Math | 1402 | 1262 |
| OTIS Mock AIME 2024-2025 | — | 8.5% |
| Omni-MATH | — | 28.1% |
| LiveBench Math | — | 42.5% |
| MATH Level 5 | — | 50.3% |
| FrontierMath (Feb 2025 set) | — | 0.3% |
Knowledge Not comparable
DeepSeek-V3.1-Terminus: —, Mistral Large: 30.1 (#230)
| Benchmark | DeepSeek-V3.1-Terminus | Mistral Large |
|---|---|---|
| GPQA Diamond | — | 51.3% |
| MMLU-Pro | — | 59.9% |
| Confabulations | — | 21.4% |
| Vectara Hallucination Rate | — | 4.5% |
| GPQA (HELM) | — | 43.5% |
| LMArena Expert | — | 1232 |
| MMLU | — | 80% |
Multilingual DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 52.1 (#92), Mistral Large: 40.0 (#219)
| Benchmark | DeepSeek-V3.1-Terminus | Mistral Large |
|---|---|---|
| LMArena Non-English | 1407 | 1237 |
| LMArena Russian | 1436 | 1257 |
| LMArena Chinese | — | 1240 |
| LMArena French | — | 1325 |
| LMArena German | — | 1254 |
| LMArena Japanese | — | 1188 |
| LMArena Korean | — | 1202 |
| LMArena Spanish | — | 1268 |
Instruction Following DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 74.0 (#106), Mistral Large: 67.9 (#191)
| Benchmark | DeepSeek-V3.1-Terminus | Mistral Large |
|---|---|---|
| LMArena Instruction Following | 1404 | 1249 |
| LiveBench Instruction Following | — | 67.9% |
| IFEval | — | 87.7% |
Long Context DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 43.4 (#97), Mistral Large: 38.3 (#199)
| Benchmark | DeepSeek-V3.1-Terminus | Mistral Large |
|---|---|---|
| LMArena Longer Query | 1421 | 1261 |
Writing & Preference DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 61.0 (#92), Mistral Large: 40.7 (#242)
| Benchmark | DeepSeek-V3.1-Terminus | Mistral Large |
|---|---|---|
| LMArena Text | 1419 | 1266 |
| LMArena Creative Writing | 1403 | 1243 |
| LMArena Multi-Turn | 1411 | 1260 |
| Short-Story Creative Writing | — | 69% |
| EQ-Bench Creative Writing | — | 985 |
| WildBench | — | 80.1% |
| LiveBench Language | — | 39.4% |
Frequently asked questions
Is DeepSeek-V3.1-Terminus better than Mistral Large?
DeepSeek-V3.1-Terminus is the stronger model overall, scoring 43.1 to 31.9 on the Noometry Index.
Which is cheaper, DeepSeek-V3.1-Terminus or Mistral Large?
DeepSeek-V3.1-Terminus is cheaper. It lists at $0.27 per million input tokens and $1 per million output tokens; Mistral Large lists at $2 and $6.
Is DeepSeek-V3.1-Terminus or Mistral Large better for coding?
DeepSeek-V3.1-Terminus scores higher on coding benchmarks: 42.0 versus 34.3 in the Noometry coding category.
Which has the bigger context window?
DeepSeek-V3.1-Terminus does, with 164K tokens against 131K.
How many benchmarks do DeepSeek-V3.1-Terminus and Mistral Large share?
15 benchmarks have published results for both models. DeepSeek-V3.1-Terminus has 16 scored results on Noometry and Mistral Large has 51.