Model comparison
DeepSeek V4 Pro vs o1-mini
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 34.0 on the Noometry Index.
Last verified . 23 shared benchmarks.
DeepSeekSummary
- They share 23 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 9 categories and o1-mini in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 8.8.
- The biggest single-benchmark swing is ARC-AGI-1: 90.5% for DeepSeek V4 Pro and 14% for o1-mini.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4 Pro | o1-mini | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 54.3 | 34.0 |
| Released | 2026-04-24 | 2024-09-12 |
| Weights | Open | Proprietary |
| Context window | 1M | — |
| Max output | 393K | — |
| Input $ / M tokens | $0.66 | — |
| Output $ / M tokens | $1.98 | — |
| Results tracked | 48 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
DeepSeek V4 Pro: 52.4 (#34), o1-mini: 35.5 (#224)
| Benchmark | DeepSeek V4 Pro | o1-mini |
|---|---|---|
| WeirdML | 66.2% | 36.3% |
| LMArena Coding | 1470 | 1362 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| Aider Polyglot | — | 32.9% |
| LMArena WebDev | 1582 | — |
| SciCode | 51% | — |
| LiveBench Coding | — | 48% |
| ALE-Bench | 1,403 | — |
| HumanEval+ | — | 89% |
| MBPP+ | — | 78.8% |
Agentic & Tool Use DeepSeek V4 Pro leads
DeepSeek V4 Pro: 32.8 (#58), o1-mini: 24.6 (#118)
| Benchmark | DeepSeek V4 Pro | o1-mini |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Cybench | — | 10% |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), o1-mini: 8.8 (#346)
| Benchmark | DeepSeek V4 Pro | o1-mini |
|---|---|---|
| ARC-AGI-2 | 61.3% | 0.8% |
| ARC-AGI-1 | 90.5% | 14% |
| LMArena Hard Prompts | 1461 | 1333 |
| Epoch Capabilities Index | 155.31 | 135.82 |
| SimpleBench | — | 18.1% |
| Kagi LLM Benchmark | 53.5% | — |
| NYT Connections (extended) | 91.3% | — |
| CritPt | 18% | — |
| Chess Puzzles | 47% | — |
| LiveBench Reasoning | — | 72.3% |
| Mystery Game Puzzles | 43% | — |
| DTBench | 93.9% | — |
| LiveBench Data Analysis | — | 57.9% |
| LMCA | 45.5% | — |
| Surface Evolver Bench | 40% | — |
| ForecastBench | 56.1 | — |
| LiveBench | — | 57.8% |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), o1-mini: 35.4 (#186)
| Benchmark | DeepSeek V4 Pro | o1-mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.6% | 46.9% |
| LMArena Math | 1455 | 1358 |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| ProofBench | 50% | — |
| LiveBench Math | — | 62% |
| MATH Level 5 | — | 89.2% |
| FrontierMath (Feb 2025 set) | — | 1.7% |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), o1-mini: 34.9 (#192)
| Benchmark | DeepSeek V4 Pro | o1-mini |
|---|---|---|
| GPQA Diamond | 91.7% | 62.4% |
| LMArena Expert | 1464 | 1316 |
| SimpleQA Verified | 52.9% | — |
| Confabulations | — | 18.6% |
| Vectara Hallucination Rate | 8.6% | — |
Multilingual DeepSeek V4 Pro leads
DeepSeek V4 Pro: 54.4 (#45), o1-mini: 43.6 (#182)
| Benchmark | DeepSeek V4 Pro | o1-mini |
|---|---|---|
| LMArena Non-English | 1439 | 1289 |
| LMArena Chinese | 1486 | 1314 |
| LMArena French | 1472 | 1293 |
| LMArena German | 1458 | 1278 |
| LMArena Japanese | 1445 | 1245 |
| LMArena Korean | 1447 | 1223 |
| LMArena Russian | 1453 | 1283 |
| LMArena Spanish | 1458 | 1303 |
Instruction Following DeepSeek V4 Pro leads
DeepSeek V4 Pro: 76.1 (#47), o1-mini: 66.7 (#206)
| Benchmark | DeepSeek V4 Pro | o1-mini |
|---|---|---|
| LMArena Instruction Following | 1448 | 1304 |
| LiveBench Instruction Following | — | 65.4% |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#51), o1-mini: 40.1 (#161)
| Benchmark | DeepSeek V4 Pro | o1-mini |
|---|---|---|
| LMArena Longer Query | 1458 | 1320 |
| CL-bench Life | 13.5% | — |
Writing & Preference DeepSeek V4 Pro leads
DeepSeek V4 Pro: 65.5 (#46), o1-mini: 48.4 (#202)
| Benchmark | DeepSeek V4 Pro | o1-mini |
|---|---|---|
| LMArena Text | 1451 | 1317 |
| LMArena Creative Writing | 1446 | 1244 |
| LMArena Multi-Turn | 1467 | 1314 |
| Short-Story Creative Writing | — | 64.9% |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
| LiveBench Language | — | 40.9% |
Frequently asked questions
Is DeepSeek V4 Pro better than o1-mini?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 34.0 on the Noometry Index.
Is DeepSeek V4 Pro or o1-mini better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 35.5 in the Noometry coding category.
How many benchmarks do DeepSeek V4 Pro and o1-mini share?
23 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and o1-mini has 39.