Llama 4 Maverick 17B 128E Instruct Logo

Llama 4 Maverick 17B 128E Instruct

Provider: Azure_ai

A natively multimodal, instruction-tuned model featuring a 17 B activated parameter MoE architecture (128 experts, ~400 B total). Processes text + image inputs and produces text output.

Key Information:

  • Identifier: llama-4-maverick-17b-128e-instruct-2025-04-05
  • Fine-tunable: Yes
  • Standard Model: No