Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16
Released by RedHatAI in 2025, Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 is a 17 billion parameter image-generation model. Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 is an open-weights image model with roughly 17 billion parameters.
by RedHatAI · 17B parameters
Best for
Ways to use Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 in osFoundry
Connect with your own key (BYOK)
Open the key dialog and paste your RedHatAI API key. osFoundry discovers Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 automatically — assign it to a Maestro role (router, direct, orchestrator, or fallback) in the Pipeline tab and it is live in every chat. Your key, your provider account — no token markup.
Deploy a dedicated endpoint
Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 is open-weights — run it locally for free, or deploy a dedicated GPU endpoint in your workspace for reserved capacity with no rate limits.
Use it in a Room App
Room Apps declare AI features in their manifest, then call them with invokeAI:
import { invokeAI } from '@osfoundry/app-sdk'
// 'summarize' is an AI feature declared in your app manifest.
const result = await invokeAI('summarize', userText)
Call it from your own apps
Once a model is wired into your workspace you can host it as an API and reach it from your own services, scripts, or CI — outside osFoundry.
What hardware can run Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16
Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 runs on a single 16GB consumer GPU (~11 GB VRAM with KV-cache headroom). Full-precision inference fits on a single H100 80GB at FP16 precision (~41 GB).
Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 vs similar models
Licence
Unspecified — Licence terms not specified — verify the upstream model card before commercial use.
Check upstream documentation.
Frequently asked about Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16
Is Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 free to use?
Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 is free to run locally on your own hardware. Hosted access through osFoundry is metered (input Free (local), output Free (local)). You can switch between local and hosted at any time.
Can I use Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 commercially?
Commercial use is allowed with conditions. Licence terms not specified — verify the upstream model card before commercial use. Check upstream documentation.
How much VRAM does Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 need?
Approximately 11 GB at Q4 quantisation, or 41 GB at full FP16 precision. Fits on a single 24GB consumer GPU.
Can I run Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 locally?
Yes. Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 is open-weights and runs locally on a workstation GPU. osFoundry's local runtime handles model loading, quantisation, and routing.
What is Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 best at?
Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 is well-suited to image text to text.
How do I use Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 in osFoundry?
Paste your RedHatAI API key in the key dialog (or deploy the open weights for self-hostable models), assign Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16 to a Maestro role in the Pipeline tab, then use it in chat, Room Apps via invokeAI, or your own apps.
Published by RedHatAI on June 12, 2025. Source: https://huggingface.co/RedHatAI/Llama-4-Maverick-17B-128E-Instruct-quantized.w4a16