science_1bmix_m32-e52b113b-not_easy_1e-4_1500
graf's science_1bmix_m32-e52b113b-not_easy_1e-4_1500 packs 113 billion parameters into a chat model. science_1bmix_m32-e52b113b-not_easy_1e-4_1500 is an open-weights chat model with roughly 113 billion parameters.
by graf · 113B parameters
Best for
- complex multi-step reasoning
- agent orchestration with tool use
- long-document analysis and summarisation
Ways to use science_1bmix_m32-e52b113b-not_easy_1e-4_1500 in osFoundry
Connect with your own key (BYOK)
Open the key dialog and paste your graf API key. osFoundry discovers science_1bmix_m32-e52b113b-not_easy_1e-4_1500 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
science_1bmix_m32-e52b113b-not_easy_1e-4_1500 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 science_1bmix_m32-e52b113b-not_easy_1e-4_1500
science_1bmix_m32-e52b113b-not_easy_1e-4_1500 runs on a single A100 80GB or H100 80GB at Q4 quantisation (~68 GB VRAM with KV-cache headroom). Full-precision inference requires multiple H100/H200 GPUs at FP16 (~272 GB).
science_1bmix_m32-e52b113b-not_easy_1e-4_1500 vs similar models
Licence
Unspecified — Licence terms not specified — verify the upstream model card before commercial use.
Check upstream documentation.
Frequently asked about science_1bmix_m32-e52b113b-not_easy_1e-4_1500
Is science_1bmix_m32-e52b113b-not_easy_1e-4_1500 free to use?
science_1bmix_m32-e52b113b-not_easy_1e-4_1500 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 science_1bmix_m32-e52b113b-not_easy_1e-4_1500 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 science_1bmix_m32-e52b113b-not_easy_1e-4_1500 need?
Approximately 68 GB at Q4 quantisation, or 272 GB at full FP16 precision. Fits on a single A100/H100 80GB.
Can I run science_1bmix_m32-e52b113b-not_easy_1e-4_1500 locally?
Yes. science_1bmix_m32-e52b113b-not_easy_1e-4_1500 is open-weights and runs locally on a workstation GPU. osFoundry's local runtime handles model loading, quantisation, and routing.
What is science_1bmix_m32-e52b113b-not_easy_1e-4_1500 best at?
science_1bmix_m32-e52b113b-not_easy_1e-4_1500 is well-suited to complex multi-step reasoning, agent orchestration with tool use, long-document analysis and summarisation.
How do I use science_1bmix_m32-e52b113b-not_easy_1e-4_1500 in osFoundry?
Paste your graf API key in the key dialog (or deploy the open weights for self-hostable models), assign science_1bmix_m32-e52b113b-not_easy_1e-4_1500 to a Maestro role in the Pipeline tab, then use it in chat, Room Apps via invokeAI, or your own apps.
Published by graf on April 25, 2026. Source: https://huggingface.co/graf/science_1bmix_m32-e52b113b-not_easy_1e-4_1500