Kimi-K2.5-Pruned-20B
Built by ubicloud, Kimi-K2.5-Pruned-20B is a 20 billion parameter chat model. Kimi-K2.5-Pruned-20B is an open-weights chat model with roughly 20 billion parameters.
by ubicloud · 20B parameters
Best for
- low-latency chat and routing
- request routing and triage
- text classification
Ways to use Kimi-K2.5-Pruned-20B in osFoundry
Connect with your own key (BYOK)
Open the key dialog and paste your ubicloud API key. osFoundry discovers Kimi-K2.5-Pruned-20B 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
Kimi-K2.5-Pruned-20B 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 Kimi-K2.5-Pruned-20B
Kimi-K2.5-Pruned-20B runs on a single 16GB consumer GPU (~12 GB VRAM with KV-cache headroom). Full-precision inference fits on a single H100 80GB at FP16 precision (~48 GB).
Kimi-K2.5-Pruned-20B vs similar models
Licence
Unspecified — Licence terms not specified — verify the upstream model card before commercial use.
Check upstream documentation.
Frequently asked about Kimi-K2.5-Pruned-20B
Is Kimi-K2.5-Pruned-20B free to use?
Kimi-K2.5-Pruned-20B 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 Kimi-K2.5-Pruned-20B 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 Kimi-K2.5-Pruned-20B need?
Approximately 12 GB at Q4 quantisation, or 48 GB at full FP16 precision. Fits on a single 24GB consumer GPU.
Can I run Kimi-K2.5-Pruned-20B locally?
Yes. Kimi-K2.5-Pruned-20B is open-weights and runs locally on a workstation GPU. osFoundry's local runtime handles model loading, quantisation, and routing.
What is Kimi-K2.5-Pruned-20B best at?
Kimi-K2.5-Pruned-20B is well-suited to low-latency chat and routing, request routing and triage, text classification.
How do I use Kimi-K2.5-Pruned-20B in osFoundry?
Paste your ubicloud API key in the key dialog (or deploy the open weights for self-hostable models), assign Kimi-K2.5-Pruned-20B to a Maestro role in the Pipeline tab, then use it in chat, Room Apps via invokeAI, or your own apps.
Published by ubicloud on March 20, 2026. Source: https://huggingface.co/ubicloud/Kimi-K2.5-Pruned-20B