Liquid AI LFM 2.5 Is Now Available in Puter.js

Puter.js now supports LFM 2.5, Liquid AI's next-generation family of edge-optimized models designed for real on-device AI agents.

What is LFM 2.5?

LFM 2.5 is Liquid AI's most capable release for edge AI deployment. Built on their hybrid architecture that combines grouped query attention with short convolutional layers, these models deliver best-in-class results while maintaining blazing inference speed.

Two variants are available:

• LFM 2.5 Instruct: General-purpose model optimized for agentic tasks, data extraction, and RAG
• LFM 2.5 Thinking: Reasoning-enhanced variant that uses chain-of-thought while requiring fewer output tokens than comparable thinking models

Examples

Basic Chat

puter.ai.chat("Explain the benefits of edge computing for AI applications", {
    model: "liquid/lfm-2.5-1.2b-instruct:free"
})

Chain-of-Thought Reasoning

puter.ai.chat("A store sells apples for $2 each and oranges for $3 each. If I buy 5 fruits and spend exactly $12, how many of each fruit did I buy?",
    { model: "liquid/lfm-2.5-1.2b-thinking:free" }
)

Data Extraction

puter.ai.chat(`Extract the product name, price, and availability from this text:
"The new XPS 15 laptop is now available for $1,299. Currently in stock with free shipping."`,
    { model: "liquid/lfm-2.5-1.2b-instruct:free" }
)

Get Started Now

Just add one script tag to your HTML:

<script src="https://js.puter.com/v2/"></script>

No API keys or account needed. Start building with LFM 2.5 immediately.

Learn more:

• Free, Unlimited Liquid AI API
• Puter.js AI Documentation