Deploy local agents everywhere with LFM2.5-2.6B
A Blog post by Liquid AI on Hugging Face Back to Articles a]:hidden"> Deploy local agents everywhere with LFM2. 6B Team Article Published August 4, 2026 Upvote 5 Leonie Monigatti iamleonie Follow LiquidAI Sergei Tilga tilgasergey Follow LiquidAI Sinoué GAD GAD-cell Follow LiquidAI Song Duong sduong Follow LiquidAI Tim Seyde tseyde Follow LiquidAI Maxime Labonne mlabonne Follow LiquidAI LFM2. 6B is built to power capable agents entirely on-device.
Key Takeaways
- It supports tool calling and multi-step workflows while staying small and fast enough for everyday hardware, from laptops to phones.
This enables developers to deploy agents everywhere, keep data private on the device, and scale usage without a cloud inference bill.
- 6B is pre-trained on ~34T tokens, with a mid-training phase that extends the context window to 128K.
Post-training then turns the base model into an agent in four stages: Supervised fine-tuning (SFT): two rounds of SFT, weighted heavily toward agentic data like tool use, web search, and harness trajectories.
- The RL framework orchestrates the training loop by launching rollouts, collecting trajectories and rewards, and updating the model.
Actions are executed within a Sandbox Service , where the Blackbox Harness hosts the agent (e.
- It is the smallest model in the group, yet it competes with and often beats the rest.
23 For your app, the strengths are instruction following and tool use.
- 6B ships with day-one support across the inference ecosystem, including llama.
Stats & Key Facts
- #Best-in-class agent: Competitive with models 4x larger on tool use, instruction following, and multi-step agentic tasks.
- #6B against models up to ~4x its size on STEM, instruction following, tool use, and agentic tasks.
It supports tool calling and multi-step workflows while staying small and fast enough for everyday hardware, from laptops to phones. This enables developers to deploy agents everywhere, keep data private on the device, and scale usage without a cloud inference bill. Best-in-class agent: Competitive with models 4x larger on tool use, instruction following, and multi-step agentic tasks.
Agentic reinforcement learning: Trained inside the most popular agentic harnesses to improve compatibility. Efficient inference : 220 tok/s on an Apple M5 Max and 113 tok/s on an AMD Ryzen CPU, in under 2. How we built a reliable agentic model for edge devices LFM2.
6B is pre-trained on ~34T tokens, with a mid-training phase that extends the context window to 128K. Post-training then turns the base model into an agent in four stages: Supervised fine-tuning (SFT): two rounds of SFT, weighted heavily toward agentic data like tool use, web search, and harness trajectories. Teacher specialization: train one specialist teacher per domain (math, code, tool use, and more).
For more details please read the original article at Hugging Face.
Continue Learning
Comments
Sign in to join the conversation