Llama 4 Maverick Instruct

llama417B

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 pass (400B total).

Context
1.0M
Max output
8K
Input price
$0.27/1M tokens
Output price
$0.85/1M tokens

Capabilities

Reasoning & Interaction
Reasoning
Tool Calling
Output Control
Structured Output
JSON Mode
Infrastructure
Streaming
Fine-tuning

Details

Model IDmeta-llama/llama-4-maverick-17b-128e-instruct-fp8
Provider Novita AI
Creator Meta Llama
Familyllama-4
Licensellama4
Statusactive
Typechat
Input modalitiestext, image
Output modalitiestext
Open weightYes
Parameters17B
Architecture—
Context window1.0M tokens
Max input—
Max output8K tokens
Reasoning tokens—
Knowledge cutoff—
Release date
Sourceofficial

Endpoints

Chat CompletionsPOST
Generate chat responses with messageshttps://api.novita.ai/v3/openai/v1/chat/completions

Pricing

Input
$0.27
Output
$0.85
Cache write
—
Cache read
—
Batch in
—
Batch out
—

Family Comparison: llama-4

ModelContextMax outPricing
Llama 4 Scout Instruct131K131K$0.18$0.59
Llama 4 Maverick Instruct1.0M8K$0.27$0.85

API

GET/v1/models/novita/meta-llama/llama-4-maverick-17b-128e-instruct-fp8

Llama 4 Maverick Instruct

meta-llama/llama-4-maverick-17b-128e-instruct-fp8

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 pass (400B total).

Changes · 2 entries
meta-llama/llama-4-maverick-17b-128e-instruct-fp8createca7a400Nov 2, 2026, 03:02 AM
meta-llama/llama-4-maverick-17b-128e-instruct-fp8create08f0ffeMay 11, 2026, 08:27 AM