Gemini 3.8 Flash vs Grok 4.6
Specs, per-token pricing, and live benchmark scores — refreshed daily from our benchmark database.
Specs & pricing
| Gemini 3.8 Flash | Grok 4.6 | |
|---|---|---|
| Company | Google DeepMind | xAI |
| Released | Sep 2026 | Aug 2026 |
| Context window | 1.0M | 500K |
| Input /1M tokens | $0.75 | $2 |
| Output /1M tokens | $3.75 | $6 |
| Open weights | No | No |
Benchmarks
| Benchmark | Gemini 3.8 Flash | Grok 4.6 |
|---|---|---|
| GPQA | — | 93.3% |
| Speed (tok/s) | — | 51 tok/s |
Scores come from published leaderboards and official model cards (sources listed on the benchmarks page) and refresh daily. Pricing via vendor pages and OpenRouter.