Model comparison
Deepseek Coder v2 vs o1
o1 is the stronger model overall, scoring 40.9 to 35.9 on the Noometry Index.
Last verified . 19 shared benchmarks.
DeepSeekSummary
- They share 19 benchmarks with published results for both. Deepseek Coder v2 scores higher in 0 categories and o1 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where o1 leads 55.6 to 38.2.
- Deepseek Coder v2 has downloadable open weights; the other is API-only.
Side by side
| Deepseek Coder v2 | o1 | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 35.9 | 40.9 |
| Released | 2024-06-17 | 2024-09-12 |
| Weights | Open | Proprietary |
| Context window | — | 200K |
| Max output | — | 100K |
| Input $ / M tokens | — | $15 |
| Output $ / M tokens | — | $60 |
| Results tracked | 24 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o1 leads
Deepseek Coder v2: 38.1 (#183), o1: 46.1 (#70)
| Benchmark | Deepseek Coder v2 | o1 |
|---|---|---|
| LMArena Coding | 1251 | 1367 |
| HumanEval+ | 82.3% | 89% |
| MBPP+ | 75.1% | 80.2% |
| Aider Polyglot | — | 61.7% |
| WeirdML | — | 47.6% |
| BigCodeBench Instruct | 48.2% | — |
| LiveBench Coding | — | 69.7% |
| BigCodeBench Complete | 59.7% | — |
| CadEval | — | 56% |
Agentic & Tool Use Not comparable
Deepseek Coder v2: —, o1: 24.6 (#117)
| Benchmark | Deepseek Coder v2 | o1 |
|---|---|---|
| Cybench | — | 10% |
| METR Time Horizons | — | 51.1% |
Reasoning o1 leads
Deepseek Coder v2: 23.6 (#176), o1: 27.9 (#111)
| Benchmark | Deepseek Coder v2 | o1 |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1371 |
| SimpleBench | — | 41.7% |
| ARC-AGI-1 | — | 30.7% |
| Chess Puzzles | — | 15% |
| EnigmaEval | — | 5.7% |
| LiveBench Reasoning | — | 91.6% |
| DTBench | — | 74.7% |
| LiveBench Data Analysis | — | 65.5% |
| LMCA | — | 22.3% |
| Epoch Capabilities Index | — | 141.91 |
| LiveBench | — | 75.7% |
| WinoGrande | 83.7% | — |
Math o1 leads
Deepseek Coder v2: 34.9 (#190), o1: 36.1 (#175)
| Benchmark | Deepseek Coder v2 | o1 |
|---|---|---|
| LMArena Math | 1241 | 1388 |
| FrontierMath (Tiers 1-3) | — | 14.7% |
| OTIS Mock AIME 2024-2025 | — | 73.3% |
| LiveBench Math | — | 80.3% |
| MATH Level 5 | — | 94.7% |
| FrontierMath (Feb 2025 set) | — | 9.3% |
| GSM8K | 94.5% | — |
Knowledge o1 leads
Deepseek Coder v2: 32.3 (#212), o1: 41.5 (#110)
| Benchmark | Deepseek Coder v2 | o1 |
|---|---|---|
| LMArena Expert | 1181 | 1361 |
| GPQA Diamond | — | 76.8% |
| Humanity's Last Exam | — | 8% |
| SimpleQA Verified | — | 41.1% |
| Confabulations | — | 11.7% |
| ARC (AI2) Challenge | 64.3% | — |
Multimodal Not comparable
Deepseek Coder v2: —, o1: 34.2 (#93)
| Benchmark | Deepseek Coder v2 | o1 |
|---|---|---|
| LMArena Vision | — | 1168 |
| GeoBench | — | 80% |
| VPCT | — | 37% |
| SpatialViz-Bench | — | 41.4% |
Multilingual o1 leads
Deepseek Coder v2: 36.3 (#240), o1: 48.6 (#142)
| Benchmark | Deepseek Coder v2 | o1 |
|---|---|---|
| LMArena Non-English | 1182 | 1358 |
| LMArena Chinese | 1201 | 1394 |
| LMArena French | 1185 | 1344 |
| LMArena German | 1164 | 1337 |
| LMArena Japanese | 1126 | 1346 |
| LMArena Korean | 1104 | 1396 |
| LMArena Russian | 1188 | 1356 |
| LMArena Spanish | 1153 | 1345 |
Instruction Following o1 leads
Deepseek Coder v2: 61.7 (#242), o1: 74.8 (#86)
| Benchmark | Deepseek Coder v2 | o1 |
|---|---|---|
| LMArena Instruction Following | 1180 | 1367 |
| LiveBench Instruction Following | — | 81.5% |
Long Context o1 leads
Deepseek Coder v2: 37.0 (#224), o1: 50.3 (#9)
| Benchmark | Deepseek Coder v2 | o1 |
|---|---|---|
| LMArena Longer Query | 1219 | 1378 |
| Fiction.LiveBench | — | 83.3% |
Writing & Preference o1 leads
Deepseek Coder v2: 38.2 (#253), o1: 55.6 (#144)
| Benchmark | Deepseek Coder v2 | o1 |
|---|---|---|
| LMArena Text | 1191 | 1366 |
| LMArena Creative Writing | 1120 | 1348 |
| LMArena Multi-Turn | 1177 | 1369 |
| Short-Story Creative Writing | — | 70.2% |
| LiveBench Language | — | 65.4% |
Frequently asked questions
Is Deepseek Coder v2 better than o1?
o1 is the stronger model overall, scoring 40.9 to 35.9 on the Noometry Index.
Is Deepseek Coder v2 or o1 better for coding?
o1 scores higher on coding benchmarks: 46.1 versus 38.1 in the Noometry coding category.
How many benchmarks do Deepseek Coder v2 and o1 share?
19 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and o1 has 52.