Model comparison
Gemma 2 9B vs GPT-6 Astra
GPT-6 Astra is the stronger model overall, scoring 70.8 to 25.9 on the Noometry Index.
Last verified . 21 shared benchmarks.
GoogleSummary
- They share 21 benchmarks with published results for both. Gemma 2 9B scores higher in 0 categories and GPT-6 Astra in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Astra leads 93.5 to 9.9.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 0.6% for Gemma 2 9B and 100% for GPT-6 Astra.
- Gemma 2 9B has downloadable open weights; the other is API-only.
Side by side
| Gemma 2 9B | GPT-6 Astra | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 25.9 | 70.8 |
| Released | 2024-06-24 | 2026-09-03 |
| Weights | Open | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 128K |
| Input $ / M tokens | — | $10 |
| Output $ / M tokens | — | $50 |
| Results tracked | 35 | 56 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Astra leads
Gemma 2 9B: 29.4 (#304), GPT-6 Astra: 73.7 (#2)
| Benchmark | Gemma 2 9B | GPT-6 Astra |
|---|---|---|
| LMArena Coding | 1173 | 1487 |
| DeepSWE | — | 74.1% |
| FrontierCode | — | 53.3% |
| LMArena WebDev | — | 1786 |
| FrontierSWE | — | 65.5% |
| SciCode | — | 56.5% |
| GSO | — | 79.4% |
| WeirdML | — | 93.6% |
| BigCodeBench Instruct | 34.7% | — |
| LiveBench Coding | 22.5% | — |
| MirrorCode | — | 46.7% |
| BigCodeBench Complete | 40.6% | — |
| ALE-Bench | — | 2,951 |
Agentic & Tool Use Not comparable
Gemma 2 9B: —, GPT-6 Astra: 52.9 (#3)
| Benchmark | Gemma 2 9B | GPT-6 Astra |
|---|---|---|
| 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
Gemma 2 9B: 15.9 (#309), GPT-6 Astra: 85.1 (#1)
| Benchmark | Gemma 2 9B | GPT-6 Astra |
|---|---|---|
| LMArena Hard Prompts | 1171 | 1462 |
| Epoch Capabilities Index | 119.83 | 166.45 |
| ARC-AGI-2 | — | 95% |
| NYT Connections (extended) | — | 98.1% |
| ARC-AGI-1 | — | 98.5% |
| CritPt | — | 31.7% |
| Chess Puzzles | — | 72% |
| EBR-Bench | — | 76.2% |
| LiveBench Reasoning | 15.2% | — |
| Mystery Game Puzzles | — | 84% |
| DTBench | — | 97.3% |
| LiveBench Data Analysis | 36.4% | — |
| LMCA | — | 64.4% |
| Bench to the Future 3 | — | 0.14 |
| LiveBench | 28.7% | — |
| PIQA | 83.7% | — |
Math GPT-6 Astra leads
Gemma 2 9B: 9.9 (#318), GPT-6 Astra: 93.5 (#2)
| Benchmark | Gemma 2 9B | GPT-6 Astra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 0.6% | 100% |
| LMArena Math | 1183 | 1465 |
| FrontierMath (Tiers 1-3) | — | 93.7% |
| FrontierMath Tier 4 | — | 97.6% |
| ProofBench | — | 99% |
| LiveBench Math | 19.8% | — |
| MATH Level 5 | 21% | — |
| FrontierMath Erdős | — | 2.9% |
| GSM8K | 84.9% | — |
Knowledge GPT-6 Astra leads
Gemma 2 9B: 9.7 (#305), GPT-6 Astra: 75.3 (#1)
| Benchmark | Gemma 2 9B | GPT-6 Astra |
|---|---|---|
| GPQA Diamond | 27.5% | 95.8% |
| LMArena Expert | 1147 | 1483 |
| Humanity's Last Exam | — | 54.8% |
| SimpleQA Verified | — | 75.6% |
| Vectara Hallucination Rate | — | 8.7% |
| BoolQ | 85.7% | — |
| MMLU | 72.1% | — |
Multimodal Not comparable
Gemma 2 9B: —, GPT-6 Astra: 55.0 (#3)
| Benchmark | Gemma 2 9B | GPT-6 Astra |
|---|---|---|
| LMArena Vision | — | 1281 |
| Blueprint-Bench 2 | — | 49.7% |
| Furniture Assembly | — | 80% |
| LMArena Document | — | 1468 |
Multilingual GPT-6 Astra leads
Gemma 2 9B: 36.6 (#238), GPT-6 Astra: 53.7 (#61)
| Benchmark | Gemma 2 9B | GPT-6 Astra |
|---|---|---|
| LMArena Non-English | 1188 | 1430 |
| LMArena Chinese | 1185 | 1484 |
| LMArena French | 1190 | 1456 |
| LMArena German | 1186 | 1440 |
| LMArena Japanese | 1144 | 1379 |
| LMArena Korean | 1137 | 1426 |
| LMArena Russian | 1200 | 1436 |
| LMArena Spanish | 1200 | 1407 |
Instruction Following GPT-6 Astra leads
Gemma 2 9B: 57.6 (#269), GPT-6 Astra: 76.3 (#44)
| Benchmark | Gemma 2 9B | GPT-6 Astra |
|---|---|---|
| LMArena Instruction Following | 1178 | 1450 |
| LiveBench Instruction Following | 52.6% | — |
Long Context GPT-6 Astra leads
Gemma 2 9B: 36.3 (#233), GPT-6 Astra: 44.5 (#62)
| Benchmark | Gemma 2 9B | GPT-6 Astra |
|---|---|---|
| LMArena Longer Query | 1197 | 1456 |
Writing & Preference GPT-6 Astra leads
Gemma 2 9B: 32.1 (#281), GPT-6 Astra: 75.3 (#7)
| Benchmark | Gemma 2 9B | GPT-6 Astra |
|---|---|---|
| LMArena Text | 1207 | 1441 |
| LMArena Creative Writing | 1206 | 1418 |
| EQ-Bench Creative Writing | 841 | 2173 |
| LMArena Multi-Turn | 1193 | 1448 |
| LiveBench Language | 25.5% | — |
Frequently asked questions
Is Gemma 2 9B better than GPT-6 Astra?
GPT-6 Astra is the stronger model overall, scoring 70.8 to 25.9 on the Noometry Index.
Is Gemma 2 9B or GPT-6 Astra better for coding?
GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 29.4 in the Noometry coding category.
How many benchmarks do Gemma 2 9B and GPT-6 Astra share?
21 benchmarks have published results for both models. Gemma 2 9B has 35 scored results on Noometry and GPT-6 Astra has 56.