Model comparison
Deepseek Coder v2 vs GPT-5 Mini
GPT-5 Mini is the stronger model overall, scoring 41.8 to 35.9 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeekSummary
- They share 17 benchmarks with published results for both. Deepseek Coder v2 scores higher in 0 categories and GPT-5 Mini in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-5 Mini leads 55.2 to 38.2.
- Deepseek Coder v2 has downloadable open weights; the other is API-only.
Side by side
| Deepseek Coder v2 | GPT-5 Mini | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 35.9 | 41.8 |
| Released | 2024-06-17 | 2025-08-07 |
| Weights | Open | Proprietary |
| Context window | — | 400K |
| Max output | — | 128K |
| Input $ / M tokens | — | $0.25 |
| Output $ / M tokens | — | $2 |
| Results tracked | 24 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 Mini leads
Deepseek Coder v2: 38.1 (#183), GPT-5 Mini: 40.1 (#146)
| Benchmark | Deepseek Coder v2 | GPT-5 Mini |
|---|---|---|
| LMArena Coding | 1251 | 1406 |
| SWE-bench Verified | — | 64.7% |
| SWE-bench Verified (bash only) | — | 59.8% |
| SWE-bench Multilingual | — | 39.7% |
| SciCode | — | 39.2% |
| WeirdML | — | 52.7% |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| ALE-Bench | — | 799.77 |
| AlgoTune | — | 1.38 |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |
Agentic & Tool Use Not comparable
Deepseek Coder v2: —, GPT-5 Mini: 31.1 (#70)
| Benchmark | Deepseek Coder v2 | GPT-5 Mini |
|---|---|---|
| Terminal-Bench | — | 34.8% |
| Berkeley Function Calling Leaderboard | — | 55.5% |
| Vending-Bench 2 | — | -31.18 |
Reasoning Too close to call
Deepseek Coder v2: 23.6 (#176), GPT-5 Mini: 23.9 (#168)
| Benchmark | Deepseek Coder v2 | GPT-5 Mini |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1380 |
| ARC-AGI-2 | — | 4.4% |
| Kagi LLM Benchmark | — | 70.3% |
| ARC-AGI-1 | — | 54.3% |
| CritPt | — | 0% |
| Chess Puzzles | — | 30% |
| EnigmaEval | — | 8.2% |
| Mystery Game Puzzles | — | 10% |
| DTBench | — | 80.5% |
| LMCA | — | 34.2% |
| Epoch Capabilities Index | — | 145.52 |
| ForecastBench | — | 61 |
| WinoGrande | 83.7% | — |
Math GPT-5 Mini leads
Deepseek Coder v2: 34.9 (#190), GPT-5 Mini: 46.7 (#69)
| Benchmark | Deepseek Coder v2 | GPT-5 Mini |
|---|---|---|
| LMArena Math | 1241 | 1378 |
| FrontierMath (Tiers 1-3) | — | 46.7% |
| FrontierMath Tier 4 | — | 12.2% |
| OTIS Mock AIME 2024-2025 | — | 86.7% |
| ProofBench | — | 9% |
| Omni-MATH | — | 72.2% |
| MATH Level 5 | — | 97.8% |
| FrontierMath (Feb 2025 set) | — | 27.2% |
| FrontierMath Tier 4 (v1) | — | 6.3% |
| GSM8K | 94.5% | — |
Knowledge GPT-5 Mini leads
Deepseek Coder v2: 32.3 (#212), GPT-5 Mini: 45.6 (#86)
| Benchmark | Deepseek Coder v2 | GPT-5 Mini |
|---|---|---|
| LMArena Expert | 1181 | 1379 |
| GPQA Diamond | — | 75% |
| Humanity's Last Exam | — | 19.4% |
| SimpleQA Verified | — | 21.6% |
| MMLU-Pro | — | 83.5% |
| Confabulations | — | 13.3% |
| Vectara Hallucination Rate | — | 12.9% |
| GPQA (HELM) | — | 75.6% |
| ARC (AI2) Challenge | 64.3% | — |
Multimodal Not comparable
Deepseek Coder v2: —, GPT-5 Mini: 35.6 (#85)
| Benchmark | Deepseek Coder v2 | GPT-5 Mini |
|---|---|---|
| LMArena Vision | — | 1202 |
| VPCT | — | 40.2% |
Multilingual GPT-5 Mini leads
Deepseek Coder v2: 36.3 (#240), GPT-5 Mini: 48.9 (#137)
| Benchmark | Deepseek Coder v2 | GPT-5 Mini |
|---|---|---|
| LMArena Non-English | 1182 | 1363 |
| LMArena Chinese | 1201 | 1385 |
| LMArena French | 1185 | 1386 |
| LMArena German | 1164 | 1366 |
| LMArena Japanese | 1126 | 1341 |
| LMArena Korean | 1104 | 1308 |
| LMArena Russian | 1188 | 1362 |
| LMArena Spanish | 1153 | 1355 |
Instruction Following GPT-5 Mini leads
Deepseek Coder v2: 61.7 (#242), GPT-5 Mini: 76.2 (#46)
| Benchmark | Deepseek Coder v2 | GPT-5 Mini |
|---|---|---|
| LMArena Instruction Following | 1180 | 1357 |
| IFEval | — | 92.7% |
Long Context GPT-5 Mini leads
Deepseek Coder v2: 37.0 (#224), GPT-5 Mini: 41.9 (#132)
| Benchmark | Deepseek Coder v2 | GPT-5 Mini |
|---|---|---|
| LMArena Longer Query | 1219 | 1355 |
| Fiction.LiveBench | — | 69.4% |
Writing & Preference GPT-5 Mini leads
Deepseek Coder v2: 38.2 (#253), GPT-5 Mini: 55.2 (#148)
| Benchmark | Deepseek Coder v2 | GPT-5 Mini |
|---|---|---|
| LMArena Text | 1191 | 1373 |
| LMArena Creative Writing | 1120 | 1325 |
| LMArena Multi-Turn | 1177 | 1363 |
| Short-Story Creative Writing | — | 83.1% |
| EQ-Bench Creative Writing | — | 1313 |
| WildBench | — | 85.5% |
Frequently asked questions
Is Deepseek Coder v2 better than GPT-5 Mini?
GPT-5 Mini is the stronger model overall, scoring 41.8 to 35.9 on the Noometry Index.
Is Deepseek Coder v2 or GPT-5 Mini better for coding?
GPT-5 Mini scores higher on coding benchmarks: 40.1 versus 38.1 in the Noometry coding category.
How many benchmarks do Deepseek Coder v2 and GPT-5 Mini share?
17 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and GPT-5 Mini has 60.