Plain-English Summary
Meta’s larger Llama 4 model for strong multimodal assistants, coding, and enterprise applications.
Llama 4 Maverick
Meta’s larger Llama 4 model for strong multimodal assistants, coding, and enterprise applications.
Llama 4
Multimodal LLM
Open weights
Open weights: Yes
API: Via Meta and partners
Technical Notes
Native multimodal MoE model with roughly 400B total parameters, 17B active parameters, 128 experts, and up to 1M context.
Best For
Self-hosted multimodal assistants, enterprise RAG, coding
Watch Out For
Large deployment footprint; license is open-weight rather than fully open-source.
Model Specs
- Maker
- Meta
- Country
- USA
- Family
- Llama 4
- Type
- Multimodal LLM
- Release date
- 2025-04-05
- Date confidence
- Exact
- Status
- Open weights
- Context window
- 1,000,000 tokens
- Max output
- Not listed
- Parameters
- ~400B
- Active parameters
- 17B
- Architecture
- Mixture-of-experts multimodal transformer
- Reasoning
- Yes
- Tool calling
- Host / fine-tune dependent
- Structured output
- Host dependent
- API available
- Via Meta and partners
- Open weights
- Yes
- Self-hostable
- Yes
- License
- Llama 4 Community License
- Inputs
- Text, image
- Outputs
- Text
- Approx. price
- See provider pricing
- How to access
- Provider website
- Model / API ID
- Llama-4-Maverick
- Knowledge cutoff
- Not publicly disclosed
- Fine-tuning
- Fine-tuning, quantization, self-hosting
- Completeness
- 95 %
- Verified
- Jul 18, 2026
- Snapshot date
- Jul 18, 2026
- Source
- Official source