Gemma 4 31B IT vs GPT-6 Astra
Gemma 4 31B IT (Google DeepMind) and GPT-6 Astra (OpenAI) compared on benchmarks, pricing, context window, and use-case rankings.
- GPT-6 Astra leads on overall intelligence (ECI 169 vs 143).
- Gemma 4 31B IT is cheaper on input tokens — $0.13 vs $10.00 per 1M.
- GPT-6 Astra has a larger context window (1.1M vs 262K tokens).
- GPT-6 Astra ranks higher for coding (94th vs 24th percentile).
| Attribute | Gemma 4 31B IT | GPT-6 Astra |
|---|---|---|
| Scores | ||
| Intelligence (ECI) | 143 | 169 |
| Coding | 24 | 94 |
| Math | 15 | 95 |
| Reasoning & Knowledge | 17 | 99 |
| Agentic & Tools | — | 93 |
| Specifications | ||
| Developer | Google DeepMind | OpenAI |
| Family | Gemma | GPT |
| Released | Apr 2, 2026 | Sep 3, 2026 |
| Parameters | — | — |
| Availability | Open weights (restricted use) | API access |
| Context window | 262K | 1.1M |
| Price — $/M input | $0.13 | $10.00 |
| Price — $/M output | $0.38 | $50.00 |
| Inputs | image, text, video | file, image, text |
| Outputs | text | text |
| Benchmarks | ||
| AIME 2024/2025 | 73% | 100% |
| GPQA Diamond | 76% | 96% |
| SimpleQA Verified | 10% | 76% |
| WebDev Arena | 1364 | 1800 |
| WeirdML | 52% | 94% |
| APEX | — | 65% |
| ARC-AGI | — | 99% |
| ARC-AGI-2 | — | 95% |
| Humanity's Last Exam | — | 55% |
| SciCode | — | 56% |
Use-case scores are 0–100 percentile composites across each area’s benchmarks, ranked against every model from the past year. Highlighted cells lead each row. Open a model for the full picture.
Frequently asked questions
Is Gemma 4 31B IT better than GPT-6 Astra?
On Epoch AI's Capabilities Index, GPT-6 Astra scores higher (169) than Gemma 4 31B IT (143). The right pick depends on your task — compare their coding, math, and reasoning scores in the table above.
Which is cheaper, Gemma 4 31B IT or GPT-6 Astra?
Gemma 4 31B IT is cheaper on input tokens at $0.13 per million, versus $10.00 (representative OpenRouter pricing).
Which has a larger context window, Gemma 4 31B IT or GPT-6 Astra?
GPT-6 Astra supports up to 1.1M tokens, compared with 262K for the other.
Which is better for coding, Gemma 4 31B IT or GPT-6 Astra?
Across coding benchmarks like SWE-bench Verified and Terminal-Bench, GPT-6 Astra ranks higher — 94th vs 24th percentile among the models tracked on Model Beat.
Want a different match-up? Open the compare tool to add or swap models.
More comparisons
Benchmarks & model data from Epoch AI (CC BY); pricing & specs from OpenRouter. ECI = Epoch Capabilities Index.