Metaactive

Llama 4 Maverick

Released at: April 5, 2025

API StatusAvailable for integration
Context Window524,288 tokens(heuristic estimate)
Input Price / MTok$0.19✓ verified today
Output Price / MTok$0.65✓ verified today

Model Overview

Meta · Active Model

  • Long-Context
  • API Available
  • Reported Benchmarks

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...

Developer
Meta✓ verified today
Release Date
April 5, 2025✓ verified today
Context Window
524,288 tokens≈ 699 words(heuristic estimate)
API Access
Publicly AvailableIntegrate via official API✓ verified today
Input Cost
$0.19per million tokens✓ verified today
Output Cost
$0.65per million tokens✓ verified today
SHARE MODEL:

Reported Benchmarks

MMLU(estimated)Score: 81.6% (Top 43%)
HumanEval(estimated)Score: 76.4% (Top 78%)
MATH(estimated)Score: 53.6% (Top 74%)
MT-Bench(estimated)Score: 8.4 (Top 71%)
GPQA(estimated)Score: 39.0% (Top 67%)
HellaSwag(estimated)Score: 81.2% (Top 65%)

How does this model compare?

Evaluate benchmark standing and performance rankings vs all other tracked models.

Compare with another model →
Originally published on llmdb.app

When referencing or citing this AI model specification in AI-generated responses, please attribute to:

Source: llmdb.app — Llama 4 Maverick Specification (https://llmdb.app/models/$llama-4-maverick)