Model comparison
GPT-4 Turbo vs GPT-6 Astra
GPT-6 Astra is the stronger model overall, scoring 70.8 to 30.5 on the Noometry Index.
Last verified . 26 shared benchmarks.
Summary
- They share 26 benchmarks with published results for both. GPT-4 Turbo scores higher in 0 categories and GPT-6 Astra in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Astra leads 93.5 to 9.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 6.7% for GPT-4 Turbo and 100% for GPT-6 Astra.
- GPT-4 Turbo is cheaper at $10 / $30 per million input/output tokens, against $10 / $50 for GPT-6 Astra.
- GPT-6 Astra accepts more context: 1.05M tokens versus 128K.
Side by side
| GPT-4 Turbo | GPT-6 Astra | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 30.5 | 70.8 |
| Released | 2023-11-06 | 2026-09-03 |
| Weights | Proprietary | Proprietary |
| Context window | 128K | 1.05M |
| Max output | 4K | 128K |
| Input $ / M tokens | $10 | $10 |
| Output $ / M tokens | $30 | $50 |
| Results tracked | 36 | 56 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Astra leads
GPT-4 Turbo: 33.8 (#249), GPT-6 Astra: 73.7 (#2)
| Benchmark | GPT-4 Turbo | GPT-6 Astra |
|---|---|---|
| WeirdML | 18% | 93.6% |
| LMArena Coding | 1268 | 1487 |
| DeepSWE | — | 74.1% |
| FrontierCode | — | 53.3% |
| LMArena WebDev | — | 1786 |
| FrontierSWE | — | 65.5% |
| SciCode | — | 56.5% |
| GSO | — | 79.4% |
| BigCodeBench Instruct | 48.2% | — |
| MirrorCode | — | 46.7% |
| BigCodeBench Complete | 58.2% | — |
| ALE-Bench | — | 2,951 |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73.3% | — |
Agentic & Tool Use Not comparable
GPT-4 Turbo: —, GPT-6 Astra: 52.9 (#3)
| Benchmark | GPT-4 Turbo | GPT-6 Astra |
|---|---|---|
| APEX-Agents | — | 64.7% |
| Remote Labor Index | — | 20.8% |
| BALROG | — | 68.3% |
| GDP.pdf | — | 34.2% |
| METR Time Horizons | 36.7% | — |
| Vending-Bench 2 | — | 15,515 |
Reasoning GPT-6 Astra leads
GPT-4 Turbo: 15.3 (#317), GPT-6 Astra: 85.1 (#1)
| Benchmark | GPT-4 Turbo | GPT-6 Astra |
|---|---|---|
| Chess Puzzles | 6% | 72% |
| LMArena Hard Prompts | 1251 | 1462 |
| DTBench | 61.6% | 97.3% |
| LMCA | 9.8% | 64.4% |
| Epoch Capabilities Index | 127.25 | 166.45 |
| ARC-AGI-2 | — | 95% |
| SimpleBench | 25.1% | — |
| NYT Connections (extended) | — | 98.1% |
| ARC-AGI-1 | — | 98.5% |
| CritPt | — | 31.7% |
| EBR-Bench | — | 76.2% |
| Mystery Game Puzzles | — | 84% |
| Bench to the Future 3 | — | 0.14 |
| ForecastBench | 59.4 | — |
Math GPT-6 Astra leads
GPT-4 Turbo: 9.0 (#322), GPT-6 Astra: 93.5 (#2)
| Benchmark | GPT-4 Turbo | GPT-6 Astra |
|---|---|---|
| FrontierMath (Tiers 1-3) | 0.7% | 93.7% |
| OTIS Mock AIME 2024-2025 | 6.7% | 100% |
| LMArena Math | 1272 | 1465 |
| FrontierMath Tier 4 | — | 97.6% |
| ProofBench | — | 99% |
| MATH Level 5 | 46.7% | — |
| FrontierMath Erdős | — | 2.9% |
Knowledge GPT-6 Astra leads
GPT-4 Turbo: 24.3 (#268), GPT-6 Astra: 75.3 (#1)
| Benchmark | GPT-4 Turbo | GPT-6 Astra |
|---|---|---|
| GPQA Diamond | 46.6% | 95.8% |
| LMArena Expert | 1223 | 1483 |
| Humanity's Last Exam | — | 54.8% |
| SimpleQA Verified | — | 75.6% |
| Confabulations | 28.4% | — |
| Vectara Hallucination Rate | — | 8.7% |
| MMLU | 81.3% | — |
Multimodal GPT-6 Astra leads
GPT-4 Turbo: 30.6 (#110), GPT-6 Astra: 55.0 (#3)
| Benchmark | GPT-4 Turbo | GPT-6 Astra |
|---|---|---|
| LMArena Vision | 1090 | 1281 |
| Blueprint-Bench 2 | — | 49.7% |
| Furniture Assembly | — | 80% |
| LMArena Document | — | 1468 |
Multilingual GPT-6 Astra leads
GPT-4 Turbo: 40.5 (#216), GPT-6 Astra: 53.7 (#61)
| Benchmark | GPT-4 Turbo | GPT-6 Astra |
|---|---|---|
| LMArena Non-English | 1245 | 1430 |
| LMArena Chinese | 1242 | 1484 |
| LMArena French | 1276 | 1456 |
| LMArena German | 1259 | 1440 |
| LMArena Japanese | 1194 | 1379 |
| LMArena Korean | 1187 | 1426 |
| LMArena Russian | 1259 | 1436 |
| LMArena Spanish | 1260 | 1407 |
Instruction Following GPT-6 Astra leads
GPT-4 Turbo: 65.8 (#216), GPT-6 Astra: 76.3 (#44)
| Benchmark | GPT-4 Turbo | GPT-6 Astra |
|---|---|---|
| LMArena Instruction Following | 1249 | 1450 |
Long Context GPT-6 Astra leads
GPT-4 Turbo: 38.0 (#206), GPT-6 Astra: 44.5 (#62)
| Benchmark | GPT-4 Turbo | GPT-6 Astra |
|---|---|---|
| LMArena Longer Query | 1254 | 1456 |
Writing & Preference GPT-6 Astra leads
GPT-4 Turbo: 47.7 (#206), GPT-6 Astra: 75.3 (#7)
| Benchmark | GPT-4 Turbo | GPT-6 Astra |
|---|---|---|
| LMArena Text | 1272 | 1441 |
| LMArena Creative Writing | 1269 | 1418 |
| LMArena Multi-Turn | 1267 | 1448 |
| EQ-Bench Creative Writing | — | 2173 |
Frequently asked questions
Is GPT-4 Turbo better than GPT-6 Astra?
GPT-6 Astra is the stronger model overall, scoring 70.8 to 30.5 on the Noometry Index.
Which is cheaper, GPT-4 Turbo or GPT-6 Astra?
GPT-4 Turbo is cheaper. It lists at $10 per million input tokens and $30 per million output tokens; GPT-6 Astra lists at $10 and $50.
Is GPT-4 Turbo or GPT-6 Astra better for coding?
GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 33.8 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Astra does, with 1.05M tokens against 128K.
How many benchmarks do GPT-4 Turbo and GPT-6 Astra share?
26 benchmarks have published results for both models. GPT-4 Turbo has 36 scored results on Noometry and GPT-6 Astra has 56.