Grok 4.20
Language modelPrevious generationCard summarizedA high-performance model with industry-leading speed and agentic tool calling capabilities.
Active (legacy; superseded by Grok 4.5–4.7)● Re-checked 12 Oct 2026 · 18 facts · 0 correctionsSource: Grok 4.20 System Card (8 pp)
Use it via API
Where to call it
- xAI API
grok-4.20-0309-reasoning
Supports
- Image input
- Structured outputs
- Tool calling
- Web search
- Prompt caching
- Batch API
Good for
- General assistant: questions, reasoning, research, writing, coding, translation
- Multimodal reasoning over text and images
- Agentic tool calling (API docs)
Overview
Grok 4.20
A high-performance model with industry-leading speed and agentic tool calling capabilities.
- Cost
- $1.25 in
- Context
- 1M
- Input / output
- $1.25 / $2.50
- Max output
- —
- Knows up to
- —
- Released
- 7 Apr 2026
Grok 4.20 is a general-purpose reasoning model with single-agent and multi-agent modes. The system card covers malicious-use and loss-of-control behaviour and a dual-use capability check under xAI's Frontier AI Framework; xAI concludes that, with safeguards, it poses no significantly more risk than prior models. §1, §1.3, p.1–2
grok-4.20-0309-reasoning- Input / output per 1M
- $1.25 / $2.50
- Context window
- 1M
- Max output
- —
- Knows up to
- —
- Reasoning
- Always on (reasoning variant)
- Weights
- Closed
About this card. A model card is the maker's own account of a model: what it's for, how it was trained and tested, and where it falls short. This page condenses xAI's Grok 4.20 System Card. Every claim carries its source page.