lfm2.5

2,213 Downloads Updated 3 days ago

LFM2.5-8B-A1B, an edge model built for fast, reliable tool calling on consumer hardware.

tools thinking 8b

ollama run lfm2.5

curl http://localhost:11434/api/chat \
  -d '{
    "model": "lfm2.5",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='lfm2.5',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'lfm2.5',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Applications

Claude Code ollama launch claude --model lfm2.5

Codex App ollama launch codex-app --model lfm2.5

OpenClaw ollama launch openclaw --model lfm2.5

Hermes Agent ollama launch hermes --model lfm2.5

Codex ollama launch codex --model lfm2.5

OpenCode ollama launch opencode --model lfm2.5

Models

View all →

Name

5 models

Size / Usage

Context

Input

lfm2.5:latest

5.2GB · 125K context window · Text · 3 days ago

lfm2.5:latest

5.2GB

125K

Text

lfm2.5:8b

latest

5.2GB · 125K context window · Text · 3 days ago

lfm2.5:8b latest

5.2GB

125K

Text

Readme

LFM2.5 is a new family of hybrid models designed for on-device deployment. It builds on the LFM2 architecture with extended pre-training and reinforcement learning.

On-device personal assistant: Designed to power real-life applications, chaining tool calls, and following complex instructions on all devices.
Compressed performance: Competitive with much larger dense and MoE models on instruction following and agentic tasks.
Unmatched throughput: Fastest in its size class on both CPU and GPU inference