Llama 4 Maverick 17B Instruct

llama

Charm Hyper · llama-4-maverick-17b-128e-instruct-fp8

Capabilities

Tool CallYes

ReasoningNo

Structured OutputNo

Temperature ControlYes

AttachmentsNo

Open WeightsYes

Input

text

Output

text

Input$0.27

Output$0.90

Cache Write$0.14

Direct installation coming soon — March 2026

ProviderCharm Hyper

Model IDllama-4-maverick-17b-128e-instruct-fp8

Familyllama

Knowledge Cutoff2024-08

Release Date2026-04-30

Last Updated2026-07-22

Context430K

Max Output43K

@ai-sdk/openai-compatible