Model comparison
DeepSeek-V2.5 (Sep 2024) vs Mistral 7B
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.0 to 22.9 on the Noometry Index.
Last verified . 20 shared benchmarks.
DeepSeek
Mistral AISummary
- They share 20 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 8 categories and Mistral 7B in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V2.5 (Sep 2024) leads 34.2 to 7.3.
- The biggest single-benchmark swing is BigCodeBench Instruct: 48.6% for DeepSeek-V2.5 (Sep 2024) and 19.5% for Mistral 7B.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Mistral 7B | |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 37.0 | 22.9 |
| Released | 2024-09-06 | 2023-09-27 |
| Weights | Open | Open |
| Context window | — | 8K |
| Max output | — | 8K |
| Input $ / M tokens | — | $0.25 |
| Output $ / M tokens | — | $0.25 |
| Results tracked | 22 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 31.3 (#274), Mistral 7B: 25.9 (#321)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Mistral 7B |
|---|---|---|
| BigCodeBench Instruct | 48.6% | 19.5% |
| LMArena Coding | 1310 | 1082 |
| BigCodeBench Complete | 53.2% | 27.3% |
| HumanEval+ | 83.5% | 36% |
| MBPP+ | 74.1% | 42.1% |
| Aider Polyglot | 17.8% | — |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 24.5 (#145), Mistral 7B: 12.6 (#331)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Mistral 7B |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1067 |
| Chess Puzzles | — | 0% |
| DTBench | — | 42.5% |
| Adversarial NLI | — | 47.1% |
| BIG-Bench Hard | — | 56.1% |
| Epoch Capabilities Index | — | 111.97 |
| HellaSwag | — | 81% |
| PIQA | — | 83% |
| WinoGrande | — | 75.3% |
Math DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 33.3 (#177), Mistral 7B: 7.6 (#319)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Mistral 7B |
|---|---|---|
| LMArena Math | 1289 | 1085 |
| OTIS Mock AIME 2024-2025 | — | 0.3% |
| MATH Level 5 | — | 3.7% |
| GSM8K | — | 54.4% |
Knowledge DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 34.2 (#190), Mistral 7B: 7.3 (#306)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Mistral 7B |
|---|---|---|
| LMArena Expert | 1266 | 1036 |
| GPQA Diamond | — | 15.2% |
| ARC (AI2) Challenge | — | 78.6% |
| BoolQ | — | 87.4% |
| MMLU | — | 62.5% |
| OpenBookQA | — | 79.8% |
| TriviaQA | — | 75.2% |
Multilingual DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 42.7 (#188), Mistral 7B: 26.0 (#278)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Mistral 7B |
|---|---|---|
| LMArena Non-English | 1273 | 1012 |
| LMArena Chinese | 1317 | 1009 |
| LMArena French | 1288 | 1036 |
| LMArena German | 1258 | 987 |
| LMArena Japanese | 1228 | 878 |
| LMArena Russian | 1289 | 1018 |
| LMArena Spanish | 1248 | 1026 |
| LMArena Korean | 1209 | — |
Instruction Following DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 67.7 (#189), Mistral 7B: 54.5 (#275)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Mistral 7B |
|---|---|---|
| LMArena Instruction Following | 1280 | 1060 |
Long Context DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#169), Mistral 7B: 32.2 (#266)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Mistral 7B |
|---|---|---|
| LMArena Longer Query | 1301 | 1060 |
Writing & Preference DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 50.0 (#182), Mistral 7B: 30.9 (#281)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Mistral 7B |
|---|---|---|
| LMArena Text | 1294 | 1090 |
| LMArena Creative Writing | 1285 | 1068 |
| LMArena Multi-Turn | 1297 | 1062 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Mistral 7B?
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.0 to 22.9 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Mistral 7B better for coding?
DeepSeek-V2.5 (Sep 2024) scores higher on coding benchmarks: 31.3 versus 25.9 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Mistral 7B share?
20 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Mistral 7B has 37.