Model comparison
DeepSeek-V2.5 (Sep 2024) vs GPT-6 Astra
GPT-6 Astra is the stronger model overall, scoring 70.3 to 37.0 on the Noometry Index.
Last verified . 14 shared benchmarks.
DeepSeekSummary
- They share 14 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 0 categories and GPT-6 Astra in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Astra leads 85.1 to 24.5.
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 37.0 | 70.3 |
| Released | 2024-09-06 | 2026-09-03 |
| Weights | Open | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 128K |
| Input $ / M tokens | — | $10 |
| Output $ / M tokens | — | $50 |
| Results tracked | 22 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Astra leads
DeepSeek-V2.5 (Sep 2024): 31.3 (#274), GPT-6 Astra: 70.5 (#3)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra |
|---|---|---|
| LMArena Coding | 1310 | 1488 |
| DeepSWE | — | 74.1% |
| FrontierCode | — | 53.3% |
| Aider Polyglot | 17.8% | — |
| LMArena WebDev | — | 1792 |
| FrontierSWE | — | 65.5% |
| SciCode | — | 56.5% |
| WeirdML | — | 93.6% |
| BigCodeBench Instruct | 48.6% | — |
| MirrorCode | — | 46.7% |
| BigCodeBench Complete | 53.2% | — |
| ALE-Bench | — | 2,951 |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, GPT-6 Astra: 52.3 (#3)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra |
|---|---|---|
| APEX-Agents | — | 64.7% |
| Remote Labor Index | — | 20.8% |
| BALROG | — | 68.3% |
| Vending-Bench 2 | — | 15,515 |
Reasoning GPT-6 Astra leads
DeepSeek-V2.5 (Sep 2024): 24.5 (#145), GPT-6 Astra: 85.1 (#1)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1464 |
| ARC-AGI-2 | — | 95% |
| NYT Connections (extended) | — | 98.1% |
| ARC-AGI-1 | — | 98.5% |
| CritPt | — | 31.7% |
| Chess Puzzles | — | 72% |
| EBR-Bench | — | 76.2% |
| Mystery Game Puzzles | — | 84% |
| DTBench | — | 97.3% |
| LMCA | — | 64.4% |
| Bench to the Future 3 | — | 0.14 |
| Epoch Capabilities Index | — | 166.6 |
Math GPT-6 Astra leads
DeepSeek-V2.5 (Sep 2024): 33.3 (#177), GPT-6 Astra: 93.4 (#1)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra |
|---|---|---|
| LMArena Math | 1289 | 1463 |
| FrontierMath (Tiers 1-3) | — | 93.7% |
| FrontierMath Tier 4 | — | 97.6% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| ProofBench | — | 99% |
| FrontierMath Erdős | — | 2.9% |
Knowledge GPT-6 Astra leads
DeepSeek-V2.5 (Sep 2024): 34.2 (#190), GPT-6 Astra: 75.3 (#1)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra |
|---|---|---|
| LMArena Expert | 1266 | 1486 |
| GPQA Diamond | — | 95.8% |
| Humanity's Last Exam | — | 54.8% |
| SimpleQA Verified | — | 75.6% |
| Vectara Hallucination Rate | — | 8.7% |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, GPT-6 Astra: 55.1 (#3)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra |
|---|---|---|
| LMArena Vision | — | 1280 |
| Blueprint-Bench 2 | — | 49.7% |
| Furniture Assembly | — | 80% |
| LMArena Document | — | 1468 |
Multilingual GPT-6 Astra leads
DeepSeek-V2.5 (Sep 2024): 42.7 (#188), GPT-6 Astra: 54.0 (#53)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra |
|---|---|---|
| LMArena Non-English | 1273 | 1431 |
| LMArena Chinese | 1317 | 1495 |
| LMArena French | 1288 | 1467 |
| LMArena Russian | 1289 | 1434 |
| LMArena Spanish | 1248 | 1412 |
| LMArena German | 1258 | — |
| LMArena Japanese | 1228 | — |
| LMArena Korean | 1209 | — |
Instruction Following GPT-6 Astra leads
DeepSeek-V2.5 (Sep 2024): 67.7 (#189), GPT-6 Astra: 76.6 (#37)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra |
|---|---|---|
| LMArena Instruction Following | 1280 | 1454 |
Long Context GPT-6 Astra leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#169), GPT-6 Astra: 44.7 (#57)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra |
|---|---|---|
| LMArena Longer Query | 1301 | 1460 |
Writing & Preference GPT-6 Astra leads
DeepSeek-V2.5 (Sep 2024): 50.0 (#182), GPT-6 Astra: 75.7 (#7)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GPT-6 Astra |
|---|---|---|
| LMArena Text | 1294 | 1443 |
| LMArena Creative Writing | 1285 | 1423 |
| LMArena Multi-Turn | 1297 | 1457 |
| EQ-Bench Creative Writing | — | 2173 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than GPT-6 Astra?
GPT-6 Astra is the stronger model overall, scoring 70.3 to 37.0 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or GPT-6 Astra better for coding?
GPT-6 Astra scores higher on coding benchmarks: 70.5 versus 31.3 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and GPT-6 Astra share?
14 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and GPT-6 Astra has 51.