Model comparison
DeepSeek-V3 vs GPT-4.1 mini
DeepSeek-V3 is the stronger model overall, scoring 39.5 to 33.6 on the Noometry Index.
Last verified . 37 shared benchmarks.
DeepSeekSummary
- They share 37 benchmarks with published results for both. DeepSeek-V3 scores higher in 7 categories and GPT-4.1 mini in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in coding, where DeepSeek-V3 leads 42.3 to 30.6.
- The biggest single-benchmark swing is Aider Polyglot: 55.1% for DeepSeek-V3 and 32.4% for GPT-4.1 mini.
- DeepSeek-V3 is cheaper at $0.24 / $0.90 per million input/output tokens, against $0.40 / $1.60 for GPT-4.1 mini.
- GPT-4.1 mini accepts more context: 1.05M tokens versus 164K.
- DeepSeek-V3 has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V3 | GPT-4.1 mini | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 39.5 | 33.6 |
| Released | 2024-12-26 | 2025-04-14 |
| Weights | Open | Proprietary |
| Context window | 164K | 1.05M |
| Max output | 164K | 33K |
| Input $ / M tokens | $0.24 | $0.40 |
| Output $ / M tokens | $0.90 | $1.60 |
| Results tracked | 60 | 47 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3 leads
DeepSeek-V3: 42.3 (#106), GPT-4.1 mini: 30.6 (#293)
| Benchmark | DeepSeek-V3 | GPT-4.1 mini |
|---|---|---|
| Aider Polyglot | 55.1% | 32.4% |
| SciCode | 35.8% | 40.4% |
| WeirdML | 36.1% | 37.6% |
| BigCodeBench Instruct | 50% | 48.9% |
| LMArena Coding | 1368 | 1367 |
| SWE-bench Verified (bash only) | — | 23.9% |
| LiveBench Coding | 70.9% | — |
| BigCodeBench Complete | 62.2% | — |
| CadEval | — | 16% |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73% | — |
Agentic & Tool Use Not comparable
DeepSeek-V3: —, GPT-4.1 mini: 33.3 (#55)
| Benchmark | DeepSeek-V3 | GPT-4.1 mini |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 50.5% |
| METR Time Horizons | 49.6% | — |
Reasoning DeepSeek-V3 leads
DeepSeek-V3: 20.5 (#236), GPT-4.1 mini: 10.8 (#340)
| Benchmark | DeepSeek-V3 | GPT-4.1 mini |
|---|---|---|
| Kagi LLM Benchmark | 52.3% | 48.6% |
| CritPt | 0% | 0% |
| LMArena Hard Prompts | 1365 | 1349 |
| DTBench | 64.8% | 68.8% |
| LMCA | 15.5% | 21.1% |
| Epoch Capabilities Index | 135.94 | 135.01 |
| ARC-AGI-2 | — | 0% |
| SimpleBench | 27.2% | — |
| ARC-AGI-1 | — | 3.5% |
| Chess Puzzles | — | 7% |
| LiveBench Reasoning | 65.8% | — |
| Mystery Game Puzzles | — | 7% |
| LiveBench Data Analysis | 60.9% | — |
| BIG-Bench Hard | 87.5% | — |
| ForecastBench | 59.1 | — |
| HellaSwag | 88.9% | — |
| LiveBench | 66.9% | — |
| PIQA | 84.7% | — |
| WinoGrande | 85.2% | — |
Math DeepSeek-V3 leads
DeepSeek-V3: 32.1 (#219), GPT-4.1 mini: 24.1 (#270)
| Benchmark | DeepSeek-V3 | GPT-4.1 mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 44.7% |
| Omni-MATH | 40.3% | 49.1% |
| LMArena Math | 1373 | 1343 |
| MATH Level 5 | 75.5% | 87.3% |
| FrontierMath (Feb 2025 set) | 1.7% | 4.5% |
| FrontierMath (Tiers 1-3) | — | 6.7% |
| LiveBench Math | 73.5% | — |
Knowledge DeepSeek-V3 leads
DeepSeek-V3: 37.5 (#155), GPT-4.1 mini: 34.7 (#194)
| Benchmark | DeepSeek-V3 | GPT-4.1 mini |
|---|---|---|
| GPQA Diamond | 67.6% | 65.8% |
| MMLU-Pro | 72.3% | 78.3% |
| GPQA (HELM) | 53.8% | 61.4% |
| LMArena Expert | 1351 | 1338 |
| SimpleQA Verified | — | 12.7% |
| Confabulations | 26.1% | — |
| Vectara Hallucination Rate | 6.1% | — |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 87.2% | — |
| TriviaQA | 82.9% | — |
Multimodal Not comparable
DeepSeek-V3: —, GPT-4.1 mini: 35.8 (#82)
| Benchmark | DeepSeek-V3 | GPT-4.1 mini |
|---|---|---|
| LMArena Vision | — | 1181 |
Multilingual DeepSeek-V3 leads
DeepSeek-V3: 48.5 (#143), GPT-4.1 mini: 45.7 (#166)
| Benchmark | DeepSeek-V3 | GPT-4.1 mini |
|---|---|---|
| LMArena Non-English | 1358 | 1318 |
| LMArena Chinese | 1391 | 1329 |
| LMArena French | 1385 | 1358 |
| LMArena German | 1374 | 1351 |
| LMArena Japanese | 1333 | 1290 |
| LMArena Korean | 1319 | 1298 |
| LMArena Russian | 1373 | 1324 |
| LMArena Spanish | 1358 | 1319 |
Instruction Following Too close to call
DeepSeek-V3: 72.8 (#130), GPT-4.1 mini: 73.7 (#118)
| Benchmark | DeepSeek-V3 | GPT-4.1 mini |
|---|---|---|
| IFEval | 83.2% | 90.4% |
| LMArena Instruction Following | 1345 | 1333 |
| LiveBench Instruction Following | 81.5% | — |
Long Context DeepSeek-V3 leads
DeepSeek-V3: 34.0 (#253), GPT-4.1 mini: 31.8 (#275)
| Benchmark | DeepSeek-V3 | GPT-4.1 mini |
|---|---|---|
| Fiction.LiveBench | 50% | 44.4% |
| LMArena Longer Query | 1352 | 1344 |
Writing & Preference DeepSeek-V3 leads
DeepSeek-V3: 57.4 (#130), GPT-4.1 mini: 48.6 (#199)
| Benchmark | DeepSeek-V3 | GPT-4.1 mini |
|---|---|---|
| LMArena Text | 1375 | 1340 |
| LMArena Creative Writing | 1364 | 1300 |
| EQ-Bench Creative Writing | 1472 | 1147 |
| WildBench | 83% | 83.8% |
| LMArena Multi-Turn | 1389 | 1354 |
| Short-Story Creative Writing | 77% | — |
| LiveBench Language | 49.1% | — |
Frequently asked questions
Is DeepSeek-V3 better than GPT-4.1 mini?
DeepSeek-V3 is the stronger model overall, scoring 39.5 to 33.6 on the Noometry Index.
Which is cheaper, DeepSeek-V3 or GPT-4.1 mini?
DeepSeek-V3 is cheaper. It lists at $0.24 per million input tokens and $0.90 per million output tokens; GPT-4.1 mini lists at $0.40 and $1.60.
Is DeepSeek-V3 or GPT-4.1 mini better for coding?
DeepSeek-V3 scores higher on coding benchmarks: 42.3 versus 30.6 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 mini does, with 1.05M tokens against 164K.
How many benchmarks do DeepSeek-V3 and GPT-4.1 mini share?
37 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and GPT-4.1 mini has 47.