Peaked at #2 Now #2
Liquid AI Launches LFM2.5
Liquid AI releases LFM2.5-24B-A2B, a 24B-total, 2B-active local model with 131K context for agent tasks.
Key points
- Liquid AI releases LFM2.5-24B-A2B, the largest member of the LFM2.5 family, a reasoning model with 24B total parameters and 2B active per token.
- The model has a context expanded to 131K tokens, a 128K-token vocabulary, and 33T pretraining tokens, with open base and post-trained weights on Hugging Face.
- Liquid AI says it performs competitively with Qwen3.6-35B-A3B on agentic and tool-use tasks.
- OpenRouter says LFM2.5-24B-A2B is now available on OpenRouter.
Key points and the reaction summary are written by AI from the posts on this page. Check the original post. How we use AI
Original post
What others are saying
1 more-
OpenRouter announces Liquid AI's LFM2.5-24B-A2B is now available, a 24B MoE with 2B active parameters and reasoning, built for agents and tool use.
View on X
Top replies on X
Reaction on X: Replies were largely positive, with one user reporting running it on an RTX 3060 with full 131K context at 87 tok/s. The original poster replied with day-one support details for llama.cpp and GGUF, MLX, vLLM, SGLang and ONNX, citing 184 tokens/s decoding on Apple M5 Max.
Sign in to see 3 top replies from X
From @liquidai, @jturner, @DogukanUrker and others. Spam removed, with English and Chinese translations.
Discussion 0
Sign up Sign in to join the discussion
No comments yet. Start the conversation.