Self-hosted alternatives to PrismML
200 self-hosted apps like PrismML, closest match first.
Top picks
- O

Ollama
LaunchOllama is a tool for running open AI models on your own computer, from the terminal or through apps you build. It runs on macOS, Windows and Linux. When a model is too large for your hardware, cloud models are available too, paid for with u
Open sourceSelf-hosted
- P
Picollm
LaunchProduction-grade on-device AI SDKs for voice, language, and vision. STT, wake word, TTS, translation, OCR, and more. Runs on mobile, web, desktop, and embedded.
FreeOpen sourceSelf-hosted
- P

Prismor
LaunchSelf-hosted runtime control plane for AI agents. Observe or HITL approve or Block rogue tool calls before it executes: secret leaks, prompt injection, supply chain etc in a local dashboard. Agent agnostic (Claude, codex, langchain etc.)
FreeOpen sourceSelf-hosted
- L

LMRing
Launchlmring is a web app for comparing AI models side by side. You pick up to four models from more than 50 providers, including OpenAI, Anthropic, Google, Meta, Mistral and DeepSeek, then write one prompt and send it to all of them at once. Res
FreeOpen sourceSelf-hosted
- M

Matrixhub
LaunchMatrixHub is an open-source, self-hosted model registry for AI models, built to work with vLLM and SGLang. It is meant to be a private replacement for Hugging Face inside your own infrastructure, in the way Harbor serves Docker Hub. Its tra
FreeOpen sourceSelf-hosted
- P

PrismerCloud
LaunchPrismer Cloud is a set of APIs and services for AI agents, which it calls a knowledge drive. Its Context API fetches web content, compresses it with an LLM so it fits an agent's context window, and caches the results across a distributed ca
Open sourceSelf-hosted
More like PrismML
About PrismML

PrismML
Concentrating intelligence
PrismML builds small, efficient AI models: 1-bit and ternary models called Bonsai, which the company says pack advanced reasoning, tool use and multimodal abilities into phones, laptops, wearables and efficient data centres. Its Bonsai 2 family spans 2 billion to 27 billion parameters, with memory footprints from about 0.43 GB to 5.9 GB. The company says the family uses far less memory and energy than standard models, and it publishes benchmark comparisons for intelligence density, throughput in tokens per second and energy per token across hardware platforms. The models are available to download, and there is a demo to try. The Bonsai Image model is also listed, along with a 16-bit standard version for comparison. The company points to a whitepaper for its evaluation method. PrismML is focused on AI research into efficiency, and it has a careers page for people who want to join the team.
- Category
- AI
- Website
- prismml.com



