What is Llama 4 Maverick?

Llama 4 Maverick is cataloged under Meta / Llama with canonical ID meta/llama-4-maverick.

MEMeta/ Llama

Llama 4 Maverick

meta/llama-4-maverick

Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward...

Compare

Model information

Released

2025-04-05

Context window

Max output

16K

Model parameters

402B

Weights

Open

Modalities

Input

ImageText

Output

Text

Identity

Version: llama-4-maverick
Knowledge cutoff: 2024-08-31
License: llama4

Resources

Hugging Face

Capabilities

Open weightStructuredTemperatureTool call

Providers

Provider	Provider model ID	Protocol	Context	Max output	Capabilities	Notes	Pricing	Streaming
OpenRouter	meta-llama/llama-4-maverick	openai_chat	1M	16K	AttachmentsTool callStructuredTemperature	Hugging Face: meta-llama/Llama-4-Maverick-17B-128E-Instruct	input $0.2/M · output $0.8/M	Yes
NanoGPT	meta-llama/llama-4-maverick	openai_chat	1M	16K	AttachmentsStructured	-	input $0.15/M · output $0.6/M cache read $0.075/M	Yes
Kilo Gateway	meta-llama/llama-4-maverick	openai_chat	1M	16K	AttachmentsTool callStructuredTemperature	-	input $0.2/M · output $0.8/M	Yes
Meta	llama-4-maverick	custom	1M	16K	-	-	-	Yes