Grok 4.20
Fast reasoning model with a 2M-token context window and strict prompt adherence.
Grok 4.20 is a reasoning model with fast agentic tool calling, a low hallucination rate and strict prompt adherence, offering one of the largest context windows available.
- Provider
- xAI (SpaceXAI)
- Type
- Language model
- Released
- Mar 31, 2026
- Context window
- 2M tokens
- Max output
- 1.8M tokens
- Price
- $1.25 input / $2.50 output per 1M tokens
- Input
- text, image, file
- Output
- text
- Open weights
- No
Best for
- Long context
- Agents
- Real-time / low latency
- Extraction & classification
Strengths
- 2M-token context
- Low hallucination rate
More from xAI (SpaceXAI)
- Grok 4.7 — xAI (SpaceXAI). xAI’s flagship for long-running software engineering and self-verification.
- Grok Imagine Image 2.0 — xAI (SpaceXAI). SpaceXAI’s image model, top-5 in the image arena.
- Grok Voice Transcribe 2.0 — xAI (SpaceXAI). SpaceXAI’s low-cost, high-accuracy transcription.