DeepSeek-V3.2-REAP-345B-A37B
DeepSeek-V3.2-REAP-345B-A37B (cerebras, 2025) is a 345 billion parameter chat model. DeepSeek-V3.2-REAP-345B-A37B is an open-weights chat model with roughly 345 billion parameters.
by cerebras · 345B parameters
Best for
Ways to use DeepSeek-V3.2-REAP-345B-A37B in osFoundry
Connect with your own key (BYOK)
Open the key dialog and paste your cerebras API key. osFoundry discovers DeepSeek-V3.2-REAP-345B-A37B 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
DeepSeek-V3.2-REAP-345B-A37B 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 DeepSeek-V3.2-REAP-345B-A37B
DeepSeek-V3.2-REAP-345B-A37B runs on a multi-GPU setup or H200 141GB at Q4 (~207 GB VRAM with KV-cache headroom). Full-precision inference requires multiple H100/H200 GPUs at FP16 (~828 GB).
DeepSeek-V3.2-REAP-345B-A37B vs similar models
Licence
Unspecified — Licence terms not specified — verify the upstream model card before commercial use.
Check upstream documentation.
Frequently asked about DeepSeek-V3.2-REAP-345B-A37B
Is DeepSeek-V3.2-REAP-345B-A37B free to use?
DeepSeek-V3.2-REAP-345B-A37B 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 DeepSeek-V3.2-REAP-345B-A37B 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 DeepSeek-V3.2-REAP-345B-A37B need?
Approximately 207 GB at Q4 quantisation, or 828 GB at full FP16 precision. Requires multi-GPU at higher quantisation.
Can I run DeepSeek-V3.2-REAP-345B-A37B locally?
Yes. DeepSeek-V3.2-REAP-345B-A37B is open-weights and runs locally on a workstation GPU. osFoundry's local runtime handles model loading, quantisation, and routing.
What is DeepSeek-V3.2-REAP-345B-A37B best at?
DeepSeek-V3.2-REAP-345B-A37B is well-suited to text generation.
How do I use DeepSeek-V3.2-REAP-345B-A37B in osFoundry?
Paste your cerebras API key in the key dialog (or deploy the open weights for self-hostable models), assign DeepSeek-V3.2-REAP-345B-A37B to a Maestro role in the Pipeline tab, then use it in chat, Room Apps via invokeAI, or your own apps.
Published by cerebras on December 9, 2025. Source: https://huggingface.co/cerebras/DeepSeek-V3.2-REAP-345B-A37B