This page is part of the model documentation OpenSource Solutions keeps for the private AI assistants we build and host. We run open source models on our own servers and test them on real business questions before recommending them. All the models we test
AI model
Installed on our servers
lfm2.5:8b
Liquid AI · released 2026-05-28
Liquid AI's device-oriented mixture-of-experts model: 8.3 billion parameters, about 1.5 billion active, hybrid convolution and attention architecture, 128K context. Very fast, reasons before answering, good at tool calls.
Completion Thinking Tools
Parameters
8.5B
Q4_K_M
Memory
5.2 GB
on the graphics card
Speed
276
tokens per second on our server
Max context
128000
tokens the model can read at once
Licence and cost
- License
- LFM Open License v1.0: free under USD 10M annual revenue, commercial license above
- Price
- Free under USD 10M annual revenue; commercial license above. Runs on our servers.
- Family
- lfm2moe
Which model for your assistant?
A model alone is not an assistant. OpenSource Solutions builds the harness around it: the connection to your documents and tools, the rules it follows, the hosting on your servers or ours, and the measurement on your own questions before you choose. Your data stays on infrastructure you control.