Gemini 3.8 Live Extended Thinking vs Grok 4.20 Reasoning
Grok 4.20 Reasoning is cheaper on output tokens, while Grok 4.20 Reasoning offers a larger context window. Choose Gemini 3.8 Live Extended Thinking or Grok 4.20 Reasoning based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Gemini 3.8 Live Extended Thinking | Grok 4.20 Reasoning |
|---|---|---|
| Provider | Vercel AI Gateway | Vercel AI Gateway |
| Input / 1M tokens | $0.75 | $1.25 |
| Output / 1M tokens | $4.50 | $2.50 |
| Context window | 0 | 2M |
| Parameters | — | — |
| Open weights | No | No |
| Released | Sep 2026 | Mar 2026 |
FAQ
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.