gpt-oss-puzzle-88B
Released by nvidia in 2026, gpt-oss-puzzle-88B is a 88 billion parameter chat model. gpt-oss-puzzle-88B is an open-weights chat model with roughly 88 billion parameters.
by nvidia · 88B parameters
Best for
Ways to use gpt-oss-puzzle-88B in osFoundry
Connect with your own key (BYOK)
Open the key dialog and paste your nvidia API key. osFoundry discovers gpt-oss-puzzle-88B 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
gpt-oss-puzzle-88B 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 gpt-oss-puzzle-88B
gpt-oss-puzzle-88B runs on a single A100 80GB or H100 80GB at Q4 quantisation (~53 GB VRAM with KV-cache headroom). Full-precision inference requires multiple H100/H200 GPUs at FP16 (~212 GB).
gpt-oss-puzzle-88B vs similar models
Licence
Unspecified — Licence terms not specified — verify the upstream model card before commercial use.
Check upstream documentation.
Frequently asked about gpt-oss-puzzle-88B
Is gpt-oss-puzzle-88B free to use?
gpt-oss-puzzle-88B 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 gpt-oss-puzzle-88B 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 gpt-oss-puzzle-88B need?
Approximately 53 GB at Q4 quantisation, or 212 GB at full FP16 precision. Fits on a single A100/H100 80GB.
Can I run gpt-oss-puzzle-88B locally?
Yes. gpt-oss-puzzle-88B is open-weights and runs locally on a workstation GPU. osFoundry's local runtime handles model loading, quantisation, and routing.
What is gpt-oss-puzzle-88B best at?
gpt-oss-puzzle-88B is well-suited to text generation.
How do I use gpt-oss-puzzle-88B in osFoundry?
Paste your nvidia API key in the key dialog (or deploy the open weights for self-hostable models), assign gpt-oss-puzzle-88B to a Maestro role in the Pipeline tab, then use it in chat, Room Apps via invokeAI, or your own apps.
Published by nvidia on March 25, 2026. Source: https://huggingface.co/nvidia/gpt-oss-puzzle-88B