Llama 4 Maverick
Meta’s 128-expert open multimodal MoE model.
Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal model built on a mixture-of-experts architecture with 128 experts and 17B active parameters per forward pass.
- Provider
- Meta
- Type
- Language model
- Released
- Apr 5, 2025
- Context window
- 1M tokens
- Max output
- 16K tokens
- Price
- $0.19 input / $0.65 output per 1M tokens
- Input
- text, image
- Output
- text
- Open weights
- Yes
- License
- Llama 4 Community License
Best for
- Local & on-device
- Chat & assistants
- Vision
- Multimodal
Strengths
- Huge open ecosystem
- Fine-tuning friendly
More from Meta
- Llama 4 Scout — Meta. Efficient open model with an industry-leading 10M-token native context.
- Muse Spark 1.3 — Meta. Meta’s multimodal reasoning model for long-running agents and coding.
- Muse Image — Meta. Meta’s low-cost image model.
- V-JEPA 2 — Meta. Meta’s open JEPA world model for video understanding, prediction and robot planning.
- Llama Guard 4 12B — Meta. Open multimodal safety classifier for prompts and responses.
Tools that use it
- Amazon Bedrock — AWS. Foundation models on AWS.