Model comparison
GPT-4.5 vs GPT-6 Astra
GPT-6 Astra is the stronger model overall, scoring 70.8 to 37.2 on the Noometry Index.
Last verified . 25 shared benchmarks.
Summary
- They share 25 benchmarks with published results for both. GPT-4.5 scores higher in 0 categories and GPT-6 Astra in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Astra leads 85.1 to 13.9.
- The biggest single-benchmark swing is ARC-AGI-2: 0.8% for GPT-4.5 and 95% for GPT-6 Astra.
Side by side
| GPT-4.5 | GPT-6 Astra | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 37.2 | 70.8 |
| Released | 2025-02-27 | 2026-09-03 |
| Weights | Proprietary | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 128K |
| Input $ / M tokens | — | $10 |
| Output $ / M tokens | — | $50 |
| Results tracked | 42 | 56 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Astra leads
GPT-4.5: 42.2 (#109), GPT-6 Astra: 73.7 (#2)
| Benchmark | GPT-4.5 | GPT-6 Astra |
|---|---|---|
| WeirdML | 39.4% | 93.6% |
| LMArena Coding | 1396 | 1487 |
| DeepSWE | — | 74.1% |
| FrontierCode | — | 53.3% |
| Aider Polyglot | 44.9% | — |
| LMArena WebDev | — | 1786 |
| FrontierSWE | — | 65.5% |
| SciCode | — | 56.5% |
| GSO | — | 79.4% |
| LiveBench Coding | 75.2% | — |
| MirrorCode | — | 46.7% |
| ALE-Bench | — | 2,951 |
Agentic & Tool Use GPT-6 Astra leads
GPT-4.5: 27.9 (#97), GPT-6 Astra: 52.9 (#3)
| Benchmark | GPT-4.5 | GPT-6 Astra |
|---|---|---|
| APEX-Agents | — | 64.7% |
| Remote Labor Index | — | 20.8% |
| Cybench | 17.5% | — |
| BALROG | — | 68.3% |
| GDP.pdf | — | 34.2% |
| Vending-Bench 2 | — | 15,515 |
Reasoning GPT-6 Astra leads
GPT-4.5: 13.9 (#330), GPT-6 Astra: 85.1 (#1)
| Benchmark | GPT-4.5 | GPT-6 Astra |
|---|---|---|
| ARC-AGI-2 | 0.8% | 95% |
| ARC-AGI-1 | 10.3% | 98.5% |
| LMArena Hard Prompts | 1403 | 1462 |
| Epoch Capabilities Index | 136.74 | 166.45 |
| SimpleBench | 34.5% | — |
| NYT Connections (extended) | — | 98.1% |
| CritPt | — | 31.7% |
| Chess Puzzles | — | 72% |
| EnigmaEval | 3.2% | — |
| EBR-Bench | — | 76.2% |
| LiveBench Reasoning | 71.1% | — |
| Mystery Game Puzzles | — | 84% |
| DTBench | — | 97.3% |
| LiveBench Data Analysis | 64.3% | — |
| LMCA | — | 64.4% |
| Bench to the Future 3 | — | 0.14 |
| ForecastBench | 61.7 | — |
| LiveBench | 69% | — |
Math GPT-6 Astra leads
GPT-4.5: 32.6 (#211), GPT-6 Astra: 93.5 (#2)
| Benchmark | GPT-4.5 | GPT-6 Astra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 100% |
| LMArena Math | 1412 | 1465 |
| FrontierMath (Tiers 1-3) | — | 93.7% |
| FrontierMath Tier 4 | — | 97.6% |
| ProofBench | — | 99% |
| LiveBench Math | 69.3% | — |
| MATH Level 5 | 78.6% | — |
| FrontierMath Erdős | — | 2.9% |
Knowledge GPT-6 Astra leads
GPT-4.5: 32.5 (#211), GPT-6 Astra: 75.3 (#1)
| Benchmark | GPT-4.5 | GPT-6 Astra |
|---|---|---|
| GPQA Diamond | 68.7% | 95.8% |
| Humanity's Last Exam | 5.4% | 54.8% |
| LMArena Expert | 1394 | 1483 |
| SimpleQA Verified | — | 75.6% |
| Confabulations | 13.6% | — |
| Vectara Hallucination Rate | — | 8.7% |
Multimodal GPT-6 Astra leads
GPT-4.5: 37.6 (#71), GPT-6 Astra: 55.0 (#3)
| Benchmark | GPT-4.5 | GPT-6 Astra |
|---|---|---|
| LMArena Vision | 1195 | 1281 |
| VPCT | 45% | — |
| Blueprint-Bench 2 | — | 49.7% |
| Furniture Assembly | — | 80% |
| LMArena Document | — | 1468 |
Multilingual GPT-6 Astra leads
GPT-4.5: 52.5 (#83), GPT-6 Astra: 53.7 (#61)
| Benchmark | GPT-4.5 | GPT-6 Astra |
|---|---|---|
| LMArena Non-English | 1413 | 1430 |
| LMArena Chinese | 1421 | 1484 |
| LMArena French | 1418 | 1456 |
| LMArena German | 1457 | 1440 |
| LMArena Japanese | 1416 | 1379 |
| LMArena Korean | 1392 | 1426 |
| LMArena Russian | 1419 | 1436 |
| LMArena Spanish | — | 1407 |
Instruction Following GPT-6 Astra leads
GPT-4.5: 72.6 (#134), GPT-6 Astra: 76.3 (#44)
| Benchmark | GPT-4.5 | GPT-6 Astra |
|---|---|---|
| LMArena Instruction Following | 1404 | 1450 |
| LiveBench Instruction Following | 72.3% | — |
Long Context GPT-6 Astra leads
GPT-4.5: 40.4 (#155), GPT-6 Astra: 44.5 (#62)
| Benchmark | GPT-4.5 | GPT-6 Astra |
|---|---|---|
| LMArena Longer Query | 1406 | 1456 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference GPT-6 Astra leads
GPT-4.5: 56.9 (#134), GPT-6 Astra: 75.3 (#7)
| Benchmark | GPT-4.5 | GPT-6 Astra |
|---|---|---|
| LMArena Text | 1417 | 1441 |
| LMArena Creative Writing | 1394 | 1418 |
| EQ-Bench Creative Writing | 1258 | 2173 |
| LMArena Multi-Turn | 1444 | 1448 |
| Short-Story Creative Writing | 75.6% | — |
| LiveBench Language | 61.5% | — |
Frequently asked questions
Is GPT-4.5 better than GPT-6 Astra?
GPT-6 Astra is the stronger model overall, scoring 70.8 to 37.2 on the Noometry Index.
Is GPT-4.5 or GPT-6 Astra better for coding?
GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 42.2 in the Noometry coding category.
How many benchmarks do GPT-4.5 and GPT-6 Astra share?
25 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and GPT-6 Astra has 56.