Model comparison
DeepSeek-V2.5 (Sep 2024) vs GPT-4o mini
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 25.5 on the Noometry Index.
Last verified . 22 shared benchmarks.
DeepSeekSummary
- They share 22 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 8 categories and GPT-4o mini in 0 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek-V2.5 (Sep 2024) leads 35.9 to 10.4.
- The biggest single-benchmark swing is Aider Polyglot: 17.8% for DeepSeek-V2.5 (Sep 2024) and 3.6% for GPT-4o mini.
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V2.5 (Sep 2024) | GPT-4o mini | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 37.6 | 25.5 |
| Released | 2024-09-06 | 2024-07-18 |
| Weights | Open | Proprietary |
| Context window | — | 128K |
| Max output | — | 16K |
| Input $ / M tokens | — | $0.15 |
| Output $ / M tokens | — | $0.60 |
| Results tracked | 22 | 60 |
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): 31.7 (#281), GPT-4o mini: 22.0 (#335)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-4o mini |
|---|---|---|
| Aider Polyglot | 17.8% | 3.6% |
| BigCodeBench Instruct | 48.6% | 46.1% |
| LMArena Coding | 1309 | 1290 |
| BigCodeBench Complete | 53.2% | 57.4% |
| HumanEval+ | 83.5% | 83.5% |
| MBPP+ | 74.1% | 72.2% |
| WeirdML | — | 11.8% |
| LiveBench Coding | — | 43.1% |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, GPT-4o mini: 27.5 (#101)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-4o mini |
|---|---|---|
| BALROG | — | 17.4% |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), GPT-4o mini: 8.7 (#347)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-4o mini |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1267 |
| ARC-AGI-2 | — | 0% |
| SimpleBench | — | 10.7% |
| Kagi LLM Benchmark | — | 28.8% |
| Chess Puzzles | — | 0% |
| LiveBench Reasoning | — | 32.8% |
| Mystery Game Puzzles | — | 12% |
| DTBench | — | 54.4% |
| LiveBench Data Analysis | — | 50% |
| LMCA | — | 10.4% |
| Epoch Capabilities Index | — | 126.56 |
| LiveBench | — | 41.3% |
| PIQA | — | 88.7% |
Math DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), GPT-4o mini: 10.4 (#314)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-4o mini |
|---|---|---|
| LMArena Math | 1288 | 1267 |
| FrontierMath (Tiers 1-3) | — | 0.7% |
| OTIS Mock AIME 2024-2025 | — | 6.9% |
| Omni-MATH | — | 28% |
| LiveBench Math | — | 36.3% |
| MATH Level 5 | — | 52.6% |
| GSM8K | — | 91.3% |
Knowledge DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), GPT-4o mini: 17.7 (#284)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-4o mini |
|---|---|---|
| LMArena Expert | 1266 | 1235 |
| GPQA Diamond | — | 37.7% |
| SimpleQA Verified | — | 8.3% |
| MMLU-Pro | — | 60.3% |
| Confabulations | — | 37.2% |
| GPQA (HELM) | — | 36.8% |
| BoolQ | — | 88.7% |
| MMLU | — | 81.8% |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, GPT-4o mini: 25.9 (#122)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-4o mini |
|---|---|---|
| LMArena Vision | — | 1066 |
| Video-MME | — | 64.8% |
| GeoBench | — | 64% |
| VPCT | — | 34% |
Multilingual Too close to call
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), GPT-4o mini: 42.0 (#199)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-4o mini |
|---|---|---|
| LMArena Non-English | 1273 | 1266 |
| LMArena Chinese | 1318 | 1265 |
| LMArena French | 1289 | 1297 |
| LMArena German | 1258 | 1272 |
| LMArena Japanese | 1228 | 1216 |
| LMArena Korean | 1209 | 1195 |
| LMArena Russian | 1289 | 1275 |
| LMArena Spanish | 1248 | 1276 |
Instruction Following DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), GPT-4o mini: 61.9 (#239)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-4o mini |
|---|---|---|
| LMArena Instruction Following | 1280 | 1258 |
| LiveBench Instruction Following | — | 56.8% |
| IFEval | — | 78.2% |
Long Context Too close to call
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), GPT-4o mini: 39.1 (#186)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-4o mini |
|---|---|---|
| LMArena Longer Query | 1301 | 1289 |
Writing & Preference DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), GPT-4o mini: 39.5 (#248)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-4o mini |
|---|---|---|
| LMArena Text | 1294 | 1286 |
| LMArena Creative Writing | 1285 | 1268 |
| LMArena Multi-Turn | 1297 | 1285 |
| Short-Story Creative Writing | — | 67.2% |
| EQ-Bench Creative Writing | — | 873 |
| WildBench | — | 79.1% |
| LiveBench Language | — | 28.6% |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than GPT-4o mini?
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 25.5 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or GPT-4o mini better for coding?
DeepSeek-V2.5 (Sep 2024) scores higher on coding benchmarks: 31.7 versus 22.0 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and GPT-4o mini share?
22 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and GPT-4o mini has 60.