Model comparison
Chatgpt 4o Latest 20250326 vs DeepSeek-V3
Chatgpt 4o Latest 20250326 is the stronger model overall, scoring 43.8 to 39.5 on the Noometry Index.
Last verified . 20 shared benchmarks.
DeepSeekSummary
- They share 20 benchmarks with published results for both. Chatgpt 4o Latest 20250326 scores higher in 7 categories and DeepSeek-V3 in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Chatgpt 4o Latest 20250326 leads 33.8 to 20.5.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 75% for Chatgpt 4o Latest 20250326 and 52.3% for DeepSeek-V3.
- DeepSeek-V3 has downloadable open weights; the other is API-only.
Side by side
| Chatgpt 4o Latest 20250326 | DeepSeek-V3 | |
|---|---|---|
| Provider | OpenAI | DeepSeek |
| Noometry Index | 43.8 | 39.5 |
| Released | — | 2024-12-26 |
| Weights | Proprietary | Open |
| Context window | — | 164K |
| Max output | — | 164K |
| Input $ / M tokens | — | $0.24 |
| Output $ / M tokens | — | $0.90 |
| Results tracked | 21 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Chatgpt 4o Latest 20250326: 41.6 (#122), DeepSeek-V3: 42.3 (#106)
| Benchmark | Chatgpt 4o Latest 20250326 | DeepSeek-V3 |
|---|---|---|
| LMArena Coding | 1413 | 1368 |
| Aider Polyglot | — | 55.1% |
| SciCode | — | 35.8% |
| WeirdML | — | 36.1% |
| BigCodeBench Instruct | — | 50% |
| LiveBench Coding | — | 70.9% |
| BigCodeBench Complete | — | 62.2% |
| HumanEval+ | — | 86.6% |
| MBPP+ | — | 73% |
Agentic & Tool Use Not comparable
Chatgpt 4o Latest 20250326: —, DeepSeek-V3: —
| Benchmark | Chatgpt 4o Latest 20250326 | DeepSeek-V3 |
|---|---|---|
| METR Time Horizons | — | 49.6% |
Reasoning Chatgpt 4o Latest 20250326 leads
Chatgpt 4o Latest 20250326: 33.8 (#71), DeepSeek-V3: 20.5 (#236)
| Benchmark | Chatgpt 4o Latest 20250326 | DeepSeek-V3 |
|---|---|---|
| Kagi LLM Benchmark | 75% | 52.3% |
| LMArena Hard Prompts | 1424 | 1365 |
| SimpleBench | — | 27.2% |
| CritPt | — | 0% |
| LiveBench Reasoning | — | 65.8% |
| DTBench | — | 64.8% |
| LiveBench Data Analysis | — | 60.9% |
| LMCA | — | 15.5% |
| BIG-Bench Hard | — | 87.5% |
| Epoch Capabilities Index | — | 135.94 |
| ForecastBench | — | 59.1 |
| HellaSwag | — | 88.9% |
| LiveBench | — | 66.9% |
| PIQA | — | 84.7% |
| WinoGrande | — | 85.2% |
Math Chatgpt 4o Latest 20250326 leads
Chatgpt 4o Latest 20250326: 38.6 (#134), DeepSeek-V3: 32.1 (#219)
| Benchmark | Chatgpt 4o Latest 20250326 | DeepSeek-V3 |
|---|---|---|
| LMArena Math | 1407 | 1373 |
| OTIS Mock AIME 2024-2025 | — | 37.8% |
| Omni-MATH | — | 40.3% |
| LiveBench Math | — | 73.5% |
| MATH Level 5 | — | 75.5% |
| FrontierMath (Feb 2025 set) | — | 1.7% |
Knowledge Chatgpt 4o Latest 20250326 leads
Chatgpt 4o Latest 20250326: 39.2 (#137), DeepSeek-V3: 37.5 (#155)
| Benchmark | Chatgpt 4o Latest 20250326 | DeepSeek-V3 |
|---|---|---|
| Confabulations | 16.6% | 26.1% |
| LMArena Expert | 1401 | 1351 |
| GPQA Diamond | — | 67.6% |
| MMLU-Pro | — | 72.3% |
| Vectara Hallucination Rate | — | 6.1% |
| GPQA (HELM) | — | 53.8% |
| ARC (AI2) Challenge | — | 95.3% |
| MMLU | — | 87.2% |
| TriviaQA | — | 82.9% |
Multimodal Not comparable
Chatgpt 4o Latest 20250326: 39.6 (#58), DeepSeek-V3: —
| Benchmark | Chatgpt 4o Latest 20250326 | DeepSeek-V3 |
|---|---|---|
| LMArena Vision | 1243 | — |
Multilingual Chatgpt 4o Latest 20250326 leads
Chatgpt 4o Latest 20250326: 52.9 (#76), DeepSeek-V3: 48.5 (#143)
| Benchmark | Chatgpt 4o Latest 20250326 | DeepSeek-V3 |
|---|---|---|
| LMArena Non-English | 1419 | 1358 |
| LMArena Chinese | 1457 | 1391 |
| LMArena French | 1446 | 1385 |
| LMArena German | 1423 | 1374 |
| LMArena Japanese | 1405 | 1333 |
| LMArena Korean | 1396 | 1319 |
| LMArena Russian | 1429 | 1373 |
| LMArena Spanish | 1434 | 1358 |
Instruction Following Chatgpt 4o Latest 20250326 leads
Chatgpt 4o Latest 20250326: 74.0 (#107), DeepSeek-V3: 72.8 (#130)
| Benchmark | Chatgpt 4o Latest 20250326 | DeepSeek-V3 |
|---|---|---|
| LMArena Instruction Following | 1403 | 1345 |
| LiveBench Instruction Following | — | 81.5% |
| IFEval | — | 83.2% |
Long Context Chatgpt 4o Latest 20250326 leads
Chatgpt 4o Latest 20250326: 43.1 (#107), DeepSeek-V3: 34.0 (#253)
| Benchmark | Chatgpt 4o Latest 20250326 | DeepSeek-V3 |
|---|---|---|
| LMArena Longer Query | 1413 | 1352 |
| Fiction.LiveBench | — | 50% |
Writing & Preference Chatgpt 4o Latest 20250326 leads
Chatgpt 4o Latest 20250326: 62.6 (#72), DeepSeek-V3: 57.4 (#130)
| Benchmark | Chatgpt 4o Latest 20250326 | DeepSeek-V3 |
|---|---|---|
| LMArena Text | 1429 | 1375 |
| LMArena Creative Writing | 1405 | 1364 |
| EQ-Bench Creative Writing | 1501 | 1472 |
| LMArena Multi-Turn | 1454 | 1389 |
| Short-Story Creative Writing | — | 77% |
| WildBench | — | 83% |
| LiveBench Language | — | 49.1% |
Frequently asked questions
Is Chatgpt 4o Latest 20250326 better than DeepSeek-V3?
Chatgpt 4o Latest 20250326 is the stronger model overall, scoring 43.8 to 39.5 on the Noometry Index.
Is Chatgpt 4o Latest 20250326 or DeepSeek-V3 better for coding?
They score almost the same on coding (41.6 vs 42.3); test both on your own repository before choosing.
How many benchmarks do Chatgpt 4o Latest 20250326 and DeepSeek-V3 share?
20 benchmarks have published results for both models. Chatgpt 4o Latest 20250326 has 21 scored results on Noometry and DeepSeek-V3 has 60.