LFM2.5

Specifications

No specifications recorded yet.

Latest developments

Liquid AI Releases LFM2.5-24B-A2B: A 2B-Active MoE for On-Device Agents MarkTechPost · 15 hours ago · 33 Liquid AI Releases LFM2.5-DSpark Draft Models That Deliver Up to 3.18x Faster Decoding Without Changing Model Outputs MarkTechPost · 1 month ago · 37

LFM2.5 Q4\_0 Checkpoints from Quantization-Aware Distillation Hugging Face · 1 month ago · 36 Liquid AI Releases LFM2.5-Encoder-230M and LFM2.5-Encoder-350M: Bidirectional Encoders That Stay Fast at 8K Context on CPU MarkTechPost · 2 months ago · 33

October 2026

◆ Liquid AI releases the LFM2.5-24B-A2B open-weight reasoning model for on-device agent tool use Model release

• 14 Oct Liquid AI Releases LFM2.5-24B-A2B: A 2B-Active MoE for On-Device Agents

August 2026

◆ Liquid AI released DSpark draft model checkpoints for three LFM2.5 models, enabling speculative decoding to speed up inference Model release

• 20 Aug Liquid AI Releases LFM2.5-DSpark Draft Models That Deliver Up to 3.18x Faster Decoding Without Changing Model Outputs
• 19 Aug LFM2.5 Q4\_0 Checkpoints from Quantization-Aware Distillation

July 2026

◆ Liquid AI releases LFM2.5-Encoder models for efficient long-context CPU inference Model release

• 29 Jul Liquid AI Releases LFM2.5-Encoder-230M and LFM2.5-Encoder-350M: Bidirectional Encoders That Stay Fast at 8K Context on CPU

Relationships

Products & technology

• LFM2.5-Encoder-230M derived from this model · 1 source
• LFM2.5-Encoder-350M derived from this model · 1 source
• Liquid AI develops this model · 1 source