@liquidai: Today, we release LFM2.5-350M....
@liquidai
22 views
Mar 31, 2026
Advertisement
1
Today, we release LFM2.5-350M. Agentic loops at 350M parameters.
A 350M model trained for reliable data extraction and tool use, where models at this scale typically struggle.
<500MB when quantized, built for environments where compute, memory, and latency are constrained.
🧵
A 350M model trained for reliable data extraction and tool use, where models at this scale typically struggle.
<500MB when quantized, built for environments where compute, memory, and latency are constrained.
🧵
2
Trained on 28T tokens with scaled RL, LFM2.5-350M is a step change from LFM2-350M:
> instruction following: 18.20 → 40.69
> data extraction: 11.67 → 32.45
> tool use: 22.95 → 44.11
These are the capabilities that matter in production.
> instruction following: 18.20 → 40.69
> data extraction: 11.67 → 32.45
> tool use: 22.95 → 44.11
These are the capabilities that matter in production.
3
What matters isn’t parameter count, it’s what you can actually run:
> runs across CPUs, GPUs, and mobile
> fast, efficient, low-latency
> consistent structured outputs
> reliable function calling and agent workflows
This enables a different deployment model: move high-frequency, structured workloads closer to the data: on-device, on-prem, or at the edge instead of routing everything through a large cloud model.
> runs across CPUs, GPUs, and mobile
> fast, efficient, low-latency
> consistent structured outputs
> reliable function calling and agent workflows
This enables a different deployment model: move high-frequency, structured workloads closer to the data: on-device, on-prem, or at the edge instead of routing everything through a large cloud model.
4
Example use cases:
> document processing pipelines running locally
> lightweight agent workflows under tight latency
> systems operating offline or on-device
> document processing pipelines running locally
> lightweight agent workflows under tight latency
> systems operating offline or on-device
5
Day 0 support across the stack:
> Hardware: @AMD, @Intel, @Qualcomm
> On-device: @lmstudio , @Cactuscompute, @RunAnywhereAI , @zeticai_ , @trymirai
> Customization: @distil_labs
> Hardware: @AMD, @Intel, @Qualcomm
> On-device: @lmstudio , @Cactuscompute, @RunAnywhereAI , @zeticai_ , @trymirai
> Customization: @distil_labs
6
LFM2.5-350M: built for real workloads.
> Blog: liquid.ai/blog/lfm2-5-35…
> Weights: huggingface.co/LiquidAI/LFM2.…
> Docs: docs.liquid.ai
> Playground: playground.liquid.ai
> Blog: liquid.ai/blog/lfm2-5-35…
> Weights: huggingface.co/LiquidAI/LFM2.…
> Docs: docs.liquid.ai
> Playground: playground.liquid.ai