All Models

Llama 4 Maverick

llamaTool CallingAttachmentsOpen WeightsStructured Output

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...

Providers4
ReleasedApr 5, 2025
Input Modalitiestext, image
Output Modalitiestext
Tarsk Usecoding

Benchmarks

10
Intelligence Index
16.3
Coding Index
19.3
Math Index

Available Providers (4)

ProviderModel IDInput CostOutput CostContextMax OutputDocs
NanoGPTmeta-llama/llama-4-maverick$0.15/MTok$0.60/MTok1.0M65.5K
Kilo Gatewaymeta-llama/llama-4-maverick$0.19/MTok$0.65/MTok128K16.4K
OpenRoutermeta-llama/llama-4-maverick$0.19/MTok$0.65/MTok1.0M16.4K
DigitalOceanllama-4-maverick$0.25/MTok$0.87/MTok128K16.4K

Capabilities

Reasoning
Tool Calling
Attachments
Open Weights
Structured Output