K-EXAONE-236B-A23B-MLX-6.5bit
inferencerlabs's K-EXAONE-236B-A23B-MLX-6.5bit packs 236 billion parameters into a chat model. K-EXAONE-236B-A23B-MLX-6.5bit is an open-weights chat model with roughly 236 billion parameters.
by inferencerlabs · 236B parameters
Best for
Ways to use K-EXAONE-236B-A23B-MLX-6.5bit in osFoundry
Connect with your own key (BYOK)
Open the key dialog and paste your inferencerlabs API key. osFoundry discovers K-EXAONE-236B-A23B-MLX-6.5bit 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
K-EXAONE-236B-A23B-MLX-6.5bit 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 K-EXAONE-236B-A23B-MLX-6.5bit
K-EXAONE-236B-A23B-MLX-6.5bit runs on a multi-GPU setup or H200 141GB at Q4 (~142 GB VRAM with KV-cache headroom). Full-precision inference requires multiple H100/H200 GPUs at FP16 (~567 GB).
K-EXAONE-236B-A23B-MLX-6.5bit vs similar models
Licence
Unspecified — Licence terms not specified — verify the upstream model card before commercial use.
Check upstream documentation.
Frequently asked about K-EXAONE-236B-A23B-MLX-6.5bit
Is K-EXAONE-236B-A23B-MLX-6.5bit free to use?
K-EXAONE-236B-A23B-MLX-6.5bit 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 K-EXAONE-236B-A23B-MLX-6.5bit 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 K-EXAONE-236B-A23B-MLX-6.5bit need?
Approximately 142 GB at Q4 quantisation, or 567 GB at full FP16 precision. Requires multi-GPU at higher quantisation.
Can I run K-EXAONE-236B-A23B-MLX-6.5bit locally?
Yes. K-EXAONE-236B-A23B-MLX-6.5bit is open-weights and runs locally on a workstation GPU. osFoundry's local runtime handles model loading, quantisation, and routing.
What is K-EXAONE-236B-A23B-MLX-6.5bit best at?
K-EXAONE-236B-A23B-MLX-6.5bit is well-suited to text generation.
How do I use K-EXAONE-236B-A23B-MLX-6.5bit in osFoundry?
Paste your inferencerlabs API key in the key dialog (or deploy the open weights for self-hostable models), assign K-EXAONE-236B-A23B-MLX-6.5bit to a Maestro role in the Pipeline tab, then use it in chat, Room Apps via invokeAI, or your own apps.
Published by inferencerlabs on January 3, 2026. Source: https://huggingface.co/inferencerlabs/K-EXAONE-236B-A23B-MLX-6.5bit