Model comparison
DeepSeek-V2.5 (Sep 2024) vs GPT-5.4 nano
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 37.0 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeekSummary
- They share 17 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 1 category and GPT-5.4 nano in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in coding, where GPT-5.4 nano leads 43.1 to 31.3.
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V2.5 (Sep 2024) | GPT-5.4 nano | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 37.0 | 41.9 |
| Released | 2024-09-06 | 2026-03-17 |
| Weights | Open | Proprietary |
| Context window | — | 400K |
| Max output | — | 128K |
| Input $ / M tokens | — | $0.20 |
| Output $ / M tokens | — | $1.25 |
| Results tracked | 22 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 nano leads
DeepSeek-V2.5 (Sep 2024): 31.3 (#274), GPT-5.4 nano: 43.1 (#82)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-5.4 nano |
|---|---|---|
| LMArena Coding | 1310 | 1404 |
| Aider Polyglot | 17.8% | — |
| SciCode | — | 46.9% |
| WeirdML | — | 49.2% |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| ALE-Bench | — | 1,005 |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Reasoning Too close to call
DeepSeek-V2.5 (Sep 2024): 24.5 (#145), GPT-5.4 nano: 23.8 (#159)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-5.4 nano |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1380 |
| ARC-AGI-2 | — | 5.7% |
| Kagi LLM Benchmark | — | 39.7% |
| ARC-AGI-1 | — | 51.5% |
| CritPt | — | 9.3% |
| Chess Puzzles | — | 30% |
| Mystery Game Puzzles | — | 9% |
| DTBench | — | 80.3% |
| LMCA | — | 36.9% |
| Epoch Capabilities Index | — | 145.78 |
| ForecastBench | — | 57.3 |
Math GPT-5.4 nano leads
DeepSeek-V2.5 (Sep 2024): 33.3 (#177), GPT-5.4 nano: 40.7 (#76)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-5.4 nano |
|---|---|---|
| LMArena Math | 1289 | 1406 |
| FrontierMath (Tiers 1-3) | — | 44.9% |
| FrontierMath Tier 4 | — | 12.2% |
| OTIS Mock AIME 2024-2025 | — | 87.8% |
| ProofBench | — | 5% |
| FrontierMath (Feb 2025 set) | — | 25.9% |
| FrontierMath Tier 4 (v1) | — | 6.3% |
Knowledge GPT-5.4 nano leads
DeepSeek-V2.5 (Sep 2024): 34.2 (#190), GPT-5.4 nano: 41.8 (#94)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-5.4 nano |
|---|---|---|
| LMArena Expert | 1266 | 1397 |
| GPQA Diamond | — | 78.5% |
| SimpleQA Verified | — | 11.7% |
| Vectara Hallucination Rate | — | 3.1% |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, GPT-5.4 nano: 36.0 (#73)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-5.4 nano |
|---|---|---|
| LMArena Vision | — | 1196 |
Multilingual GPT-5.4 nano leads
DeepSeek-V2.5 (Sep 2024): 42.7 (#188), GPT-5.4 nano: 48.8 (#136)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-5.4 nano |
|---|---|---|
| LMArena Non-English | 1273 | 1359 |
| LMArena Chinese | 1317 | 1391 |
| LMArena French | 1288 | 1396 |
| LMArena German | 1258 | 1368 |
| LMArena Japanese | 1228 | 1342 |
| LMArena Korean | 1209 | 1321 |
| LMArena Russian | 1289 | 1363 |
| LMArena Spanish | 1248 | 1368 |
Instruction Following GPT-5.4 nano leads
DeepSeek-V2.5 (Sep 2024): 67.7 (#189), GPT-5.4 nano: 72.1 (#139)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-5.4 nano |
|---|---|---|
| LMArena Instruction Following | 1280 | 1362 |
Long Context GPT-5.4 nano leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#169), GPT-5.4 nano: 41.6 (#131)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-5.4 nano |
|---|---|---|
| LMArena Longer Query | 1301 | 1365 |
Writing & Preference GPT-5.4 nano leads
DeepSeek-V2.5 (Sep 2024): 50.0 (#182), GPT-5.4 nano: 55.9 (#137)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-5.4 nano |
|---|---|---|
| LMArena Text | 1294 | 1372 |
| LMArena Creative Writing | 1285 | 1313 |
| LMArena Multi-Turn | 1297 | 1382 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than GPT-5.4 nano?
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 37.0 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or GPT-5.4 nano better for coding?
GPT-5.4 nano scores higher on coding benchmarks: 43.1 versus 31.3 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and GPT-5.4 nano share?
17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and GPT-5.4 nano has 40.