Model comparison
GPT-6 Astra vs Grok-2 (Dec 2024)
GPT-6 Astra is the stronger model overall, scoring 70.8 to 33.7 on the Noometry Index.
Last verified . 22 shared benchmarks.
Summary
- They share 22 benchmarks with published results for both. GPT-6 Astra scores higher in 8 categories and Grok-2 (Dec 2024) in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Astra leads 93.5 to 20.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 100% for GPT-6 Astra and 11.5% for Grok-2 (Dec 2024).
Side by side
| GPT-6 Astra | Grok-2 (Dec 2024) | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 70.8 | 33.7 |
| Released | 2026-09-03 | 2024-08-13 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | — |
| Max output | 128K | — |
| Input $ / M tokens | $10 | — |
| Output $ / M tokens | $50 | — |
| Results tracked | 56 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Astra leads
GPT-6 Astra: 73.7 (#2), Grok-2 (Dec 2024): 33.3 (#258)
| Benchmark | GPT-6 Astra | Grok-2 (Dec 2024) |
|---|---|---|
| WeirdML | 93.6% | 22.2% |
| LMArena Coding | 1487 | 1287 |
| DeepSWE | 74.1% | — |
| FrontierCode | 53.3% | — |
| LMArena WebDev | 1786 | — |
| FrontierSWE | 65.5% | — |
| SciCode | 56.5% | — |
| GSO | 79.4% | — |
| LiveBench Coding | — | 46.4% |
| MirrorCode | 46.7% | — |
| ALE-Bench | 2,951 | — |
Agentic & Tool Use Not comparable
GPT-6 Astra: 52.9 (#3), Grok-2 (Dec 2024): —
| Benchmark | GPT-6 Astra | Grok-2 (Dec 2024) |
|---|---|---|
| APEX-Agents | 64.7% | — |
| Remote Labor Index | 20.8% | — |
| BALROG | 68.3% | — |
| GDP.pdf | 34.2% | — |
| Vending-Bench 2 | 15,515 | — |
Reasoning GPT-6 Astra leads
GPT-6 Astra: 85.1 (#1), Grok-2 (Dec 2024): 16.9 (#299)
| Benchmark | GPT-6 Astra | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Hard Prompts | 1462 | 1272 |
| DTBench | 97.3% | 65.2% |
| Epoch Capabilities Index | 166.45 | 130.48 |
| ARC-AGI-2 | 95% | — |
| SimpleBench | — | 22.7% |
| NYT Connections (extended) | 98.1% | — |
| ARC-AGI-1 | 98.5% | — |
| CritPt | 31.7% | — |
| Chess Puzzles | 72% | — |
| EBR-Bench | 76.2% | — |
| LiveBench Reasoning | — | 54.8% |
| Mystery Game Puzzles | 84% | — |
| LiveBench Data Analysis | — | 54.5% |
| LMCA | 64.4% | — |
| Bench to the Future 3 | 0.14 | — |
| LiveBench | — | 54.3% |
Math GPT-6 Astra leads
GPT-6 Astra: 93.5 (#2), Grok-2 (Dec 2024): 20.8 (#284)
| Benchmark | GPT-6 Astra | Grok-2 (Dec 2024) |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 11.5% |
| LMArena Math | 1465 | 1283 |
| FrontierMath (Tiers 1-3) | 93.7% | — |
| FrontierMath Tier 4 | 97.6% | — |
| ProofBench | 99% | — |
| LiveBench Math | — | 54.9% |
| MATH Level 5 | — | 63.5% |
| FrontierMath (Feb 2025 set) | — | 0.7% |
| FrontierMath Erdős | 2.9% | — |
Knowledge GPT-6 Astra leads
GPT-6 Astra: 75.3 (#1), Grok-2 (Dec 2024): 29.8 (#233)
| Benchmark | GPT-6 Astra | Grok-2 (Dec 2024) |
|---|---|---|
| GPQA Diamond | 95.8% | 53.8% |
| LMArena Expert | 1483 | 1254 |
| Humanity's Last Exam | 54.8% | — |
| SimpleQA Verified | 75.6% | — |
| Confabulations | — | 20.1% |
| Vectara Hallucination Rate | 8.7% | — |
Multimodal Not comparable
GPT-6 Astra: 55.0 (#3), Grok-2 (Dec 2024): —
| Benchmark | GPT-6 Astra | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Vision | 1281 | — |
| Blueprint-Bench 2 | 49.7% | — |
| Furniture Assembly | 80% | — |
| LMArena Document | 1468 | — |
Multilingual GPT-6 Astra leads
GPT-6 Astra: 53.7 (#61), Grok-2 (Dec 2024): 43.1 (#188)
| Benchmark | GPT-6 Astra | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Non-English | 1430 | 1282 |
| LMArena Chinese | 1484 | 1289 |
| LMArena French | 1456 | 1318 |
| LMArena German | 1440 | 1287 |
| LMArena Japanese | 1379 | 1244 |
| LMArena Korean | 1426 | 1237 |
| LMArena Russian | 1436 | 1286 |
| LMArena Spanish | 1407 | 1281 |
Instruction Following GPT-6 Astra leads
GPT-6 Astra: 76.3 (#44), Grok-2 (Dec 2024): 66.9 (#202)
| Benchmark | GPT-6 Astra | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Instruction Following | 1450 | 1270 |
| LiveBench Instruction Following | — | 69.6% |
Long Context GPT-6 Astra leads
GPT-6 Astra: 44.5 (#62), Grok-2 (Dec 2024): 38.8 (#190)
| Benchmark | GPT-6 Astra | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Longer Query | 1456 | 1276 |
Writing & Preference GPT-6 Astra leads
GPT-6 Astra: 75.3 (#7), Grok-2 (Dec 2024): 48.6 (#198)
| Benchmark | GPT-6 Astra | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Text | 1441 | 1305 |
| LMArena Creative Writing | 1418 | 1284 |
| LMArena Multi-Turn | 1448 | 1290 |
| Short-Story Creative Writing | — | 63.6% |
| EQ-Bench Creative Writing | 2173 | — |
| LiveBench Language | — | 45.6% |
Frequently asked questions
Is GPT-6 Astra better than Grok-2 (Dec 2024)?
GPT-6 Astra is the stronger model overall, scoring 70.8 to 33.7 on the Noometry Index.
Is GPT-6 Astra or Grok-2 (Dec 2024) better for coding?
GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 33.3 in the Noometry coding category.
How many benchmarks do GPT-6 Astra and Grok-2 (Dec 2024) share?
22 benchmarks have published results for both models. GPT-6 Astra has 56 scored results on Noometry and Grok-2 (Dec 2024) has 34.