Llama 4 Maverick

Name: Llama 4 Maverick
Author: Meta

Llama 4 Maverick is a Meta model with 1M context, Not verified in / Not verified out per million tokens, last verified October 2, 2026. Open weights: yes. Public-board scores are linked to the publisher; vendor-only figures stay in the watch-out, not on the board.

Llama 4 Maverick is Meta's open-weight 400B/17B MoE with a 1M-token context and native multimodal support.

Recorded profile verification October 2, 2026Individual provider fields may change

AI model specification details
Specification	Llama 4 Maverick
Provider	Meta
Tier	Balanced
Context window	1M
Max output	Not verifiedUnverified
Input / 1M tokens	Not verifiedUnverified
Output / 1M tokens	Not verifiedUnverified
Weights	Open
Parameters	400B total / 17B active (MoE)
Reasoning levels	Not verifiedUnverified
Modalities	text, image
License	Llama 4 Community License
Released	April 5, 2025

Pricing tiers: Llama 4 Community License weights; self-hosting incurs compute costs and providers set their own tariffs. Meta's current API catalog does not list Llama 4; a first-party token tariff/endpoint is not verified. The checkpoint supports text/image input, text/code output and 1M context.

Best for

Open-weight deployments
Multimodal understanding
Self-hosting

Watch out

The custom licence requires agreement/attribution notices and Built with Llama on distributed models/products; relevant derivative-model names start Llama. Above 700M monthly active users in the month before the release date, a separate Meta grant is required. The acceptable-use policy applies. Groq lists Maverick endpoint shutdown March 9, 2026; weights and other hosts have separate availability. Hard output cap and effort levels remain unverified.

Source receipts

Catalog figures for this model were checked against the following sources.

Meta — llama-models/models/llama4/MODEL_CARD.md at main · meta-llama/llama-models (accessed 2026-10-01)
Meta — llama-models/models/llama4/LICENSE at main · meta-llama/llama-models (accessed 2026-10-01)
Meta — The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation (accessed 2026-10-01)
Groq — Model Deprecation - GroqDocs (accessed 2026-10-01)
Meta — Models (accessed 2026-10-01)

Common questions

Llama 4 Maverick

Answered from the verified figures on this page rather than general guidance.

What is Llama 4 Maverick's context window?

Llama 4 Maverick accepts about 1M tokens of context. That only matters if you routinely send very long documents, large codebases, or multi-turn histories that approach that limit.

What is Llama 4 Maverick best for?

Llama 4 Maverick is a balanced tier from Meta. It suits open-weight deployments, multimodal understanding, self-hosting. The custom licence requires agreement/attribution notices and Built with Llama on distributed models/products; relevant derivative-model names start Llama. Above 700M monthly active users in the month before the release date, a separate Meta grant is required. The acceptable-use policy applies. Groq lists Maverick endpoint shutdown March 9, 2026; weights and other hosts have separate availability. Hard output cap and effort levels remain unverified.

Can I self-host Llama 4 Maverick?

Llama 4 Maverick publishes open weights, but self-hosting depends on the licence, hardware footprint, quantisation quality, and serving stack. A hosted API is often cheaper until you have measured throughput and concurrency on your own hardware.

Model comparison catalog Benchmark library