Model comparison
DeepSeek-V3.1 vs GPT-4o mini
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 25.5 on the Noometry Index. GPT-4o mini costs 1.6× less per token, which makes it the better buy when DeepSeek-V3.1's lead doesn't matter for your workload.
Last verified . 24 shared benchmarks.
DeepSeekSummary
- They share 24 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 7 categories and GPT-4o mini in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek-V3.1 leads 38.9 to 10.4.
- The biggest single-benchmark swing is SimpleBench: 40% for DeepSeek-V3.1 and 10.7% for GPT-4o mini.
- GPT-4o mini is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.25 / $0.95 for DeepSeek-V3.1.
- DeepSeek-V3.1 accepts more context: 164K tokens versus 128K.
- DeepSeek-V3.1 has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V3.1 | GPT-4o mini | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 42.8 | 25.5 |
| Released | 2025-08-21 | 2024-07-18 |
| Weights | Open | Proprietary |
| Context window | 164K | 128K |
| Max output | 8K | 16K |
| Input $ / M tokens | $0.25 | $0.15 |
| Output $ / M tokens | $0.95 | $0.60 |
| Results tracked | 27 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.1 leads
DeepSeek-V3.1: 40.3 (#144), GPT-4o mini: 22.0 (#335)
| Benchmark | DeepSeek-V3.1 | GPT-4o mini |
|---|---|---|
| WeirdML | 38.4% | 11.8% |
| LMArena Coding | 1417 | 1290 |
| Aider Polyglot | — | 3.6% |
| BigCodeBench Instruct | — | 46.1% |
| LiveBench Coding | — | 43.1% |
| BigCodeBench Complete | — | 57.4% |
| HumanEval+ | — | 83.5% |
| MBPP+ | — | 72.2% |
Agentic & Tool Use Not comparable
DeepSeek-V3.1: —, GPT-4o mini: 27.5 (#101)
| Benchmark | DeepSeek-V3.1 | GPT-4o mini |
|---|---|---|
| BALROG | — | 17.4% |
Reasoning DeepSeek-V3.1 leads
DeepSeek-V3.1: 27.9 (#110), GPT-4o mini: 8.7 (#347)
| Benchmark | DeepSeek-V3.1 | GPT-4o mini |
|---|---|---|
| SimpleBench | 40% | 10.7% |
| Kagi LLM Benchmark | 53.2% | 28.8% |
| LMArena Hard Prompts | 1417 | 1267 |
| DTBench | 82.7% | 54.4% |
| LMCA | 24.3% | 10.4% |
| Epoch Capabilities Index | 139.92 | 126.56 |
| ARC-AGI-2 | — | 0% |
| Chess Puzzles | — | 0% |
| LiveBench Reasoning | — | 32.8% |
| Mystery Game Puzzles | — | 12% |
| LiveBench Data Analysis | — | 50% |
| ForecastBench | 58 | — |
| LiveBench | — | 41.3% |
| PIQA | — | 88.7% |
Math DeepSeek-V3.1 leads
DeepSeek-V3.1: 38.9 (#122), GPT-4o mini: 10.4 (#314)
| Benchmark | DeepSeek-V3.1 | GPT-4o mini |
|---|---|---|
| LMArena Math | 1420 | 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-V3.1 leads
DeepSeek-V3.1: 43.7 (#90), GPT-4o mini: 17.7 (#284)
| Benchmark | DeepSeek-V3.1 | GPT-4o mini |
|---|---|---|
| LMArena Expert | 1405 | 1235 |
| GPQA Diamond | — | 37.7% |
| SimpleQA Verified | — | 8.3% |
| MMLU-Pro | — | 60.3% |
| Confabulations | — | 37.2% |
| Vectara Hallucination Rate | 5.5% | — |
| GPQA (HELM) | — | 36.8% |
| BoolQ | — | 88.7% |
| MMLU | — | 81.8% |
Multimodal Not comparable
DeepSeek-V3.1: —, GPT-4o mini: 25.9 (#122)
| Benchmark | DeepSeek-V3.1 | GPT-4o mini |
|---|---|---|
| LMArena Vision | — | 1066 |
| Video-MME | — | 64.8% |
| GeoBench | — | 64% |
| VPCT | — | 34% |
Multilingual DeepSeek-V3.1 leads
DeepSeek-V3.1: 51.6 (#106), GPT-4o mini: 42.0 (#199)
| Benchmark | DeepSeek-V3.1 | GPT-4o mini |
|---|---|---|
| LMArena Non-English | 1400 | 1266 |
| LMArena Chinese | 1469 | 1265 |
| LMArena French | 1447 | 1297 |
| LMArena German | 1411 | 1272 |
| LMArena Japanese | 1378 | 1216 |
| LMArena Korean | 1337 | 1195 |
| LMArena Russian | 1405 | 1275 |
| LMArena Spanish | 1431 | 1276 |
Instruction Following DeepSeek-V3.1 leads
DeepSeek-V3.1: 73.9 (#110), GPT-4o mini: 61.9 (#239)
| Benchmark | DeepSeek-V3.1 | GPT-4o mini |
|---|---|---|
| LMArena Instruction Following | 1400 | 1258 |
| LiveBench Instruction Following | — | 56.8% |
| IFEval | — | 78.2% |
Long Context GPT-4o mini leads
DeepSeek-V3.1: 36.3 (#232), GPT-4o mini: 39.1 (#186)
| Benchmark | DeepSeek-V3.1 | GPT-4o mini |
|---|---|---|
| LMArena Longer Query | 1422 | 1289 |
| Fiction.LiveBench | 52.8% | — |
Writing & Preference DeepSeek-V3.1 leads
DeepSeek-V3.1: 60.3 (#98), GPT-4o mini: 39.5 (#248)
| Benchmark | DeepSeek-V3.1 | GPT-4o mini |
|---|---|---|
| LMArena Text | 1420 | 1286 |
| LMArena Creative Writing | 1401 | 1268 |
| EQ-Bench Creative Writing | 1436 | 873 |
| LMArena Multi-Turn | 1408 | 1285 |
| Short-Story Creative Writing | — | 67.2% |
| WildBench | — | 79.1% |
| LiveBench Language | — | 28.6% |
Frequently asked questions
Is DeepSeek-V3.1 better than GPT-4o mini?
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 25.5 on the Noometry Index. GPT-4o mini costs 1.6× less per token, which makes it the better buy when DeepSeek-V3.1's lead doesn't matter for your workload.
Which is cheaper, DeepSeek-V3.1 or GPT-4o mini?
GPT-4o mini is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; DeepSeek-V3.1 lists at $0.25 and $0.95.
Is DeepSeek-V3.1 or GPT-4o mini better for coding?
DeepSeek-V3.1 scores higher on coding benchmarks: 40.3 versus 22.0 in the Noometry coding category.
Which has the bigger context window?
DeepSeek-V3.1 does, with 164K tokens against 128K.
How many benchmarks do DeepSeek-V3.1 and GPT-4o mini share?
24 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and GPT-4o mini has 60.