# Grok 4.20 Multi-Agent API prices

Grok 4.20 Multi-Agent (grok) — 2.0M context, 2.0M max output.
7 metered per-token offers from 6 provider(s), in USD per 1M tokens.

| Provider | Input $/1M | Output $/1M | Cache read $/1M | Context |
| --- | --- | --- | --- | --- |
| OpenRouter | $1.25 | $2.5 | $0.2 | 2.0M |
| Vercel AI Gateway | $1.25 | $2.5 | $0.2 | 2.0M |
| Vercel AI Gateway | $1.25 | $2.5 | $0.2 | 2.0M |
| xAI | $1.25 | $2.5 | $0.2 | 1.0M |
| Venice AI | $1.42 | $2.83 | $0.23 | 2.0M |
| NanoGPT | $2 | $6 | $1 | 2.0M |
| Poe | $2 | $6 | $0.2 | 128k |

## Summary

- Cheapest input: $1.25 per 1M tokens (OpenRouter)
- Cheapest output: $2.5 per 1M tokens (OpenRouter)
- First-party: $2.5 per 1M output tokens (xAI)
- Free offers: none
- Subscription plans (not per-token): none
- Released: 2026-03-31
- Inputs: image, pdf, text

HTML page: https://llmprice.gitlab.io/models/grok-4-20-multi-agent/
Full dataset: https://llmprice.gitlab.io/data/catalog.json