grok-4.20-beta-0309-non-reasoning vs claude-sonnet-4-5
grok-4.20-beta-0309-non-reasoning is cheaper on output tokens, while grok-4.20-beta-0309-non-reasoning offers a larger context window. Choose grok-4.20-beta-0309-non-reasoning or claude-sonnet-4-5 based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | grok-4.20-beta-0309-non-reasoning | claude-sonnet-4-5 |
|---|---|---|
| Provider | 302.AI | 302.AI |
| Input / 1M tokens | $2.00 | $3.00 |
| Output / 1M tokens | $6.00 | $15.00 |
| Context window | 2M | 200K |
| Parameters | — | — |
| Open weights | No | No |
| Released | Mar 2026 | Sep 2025 |
FAQ
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.