In the news
LFM2.5-24B-A2B: Built for Fast On-Device Agents
Liquid AI · Published · 3 min read
In 30 seconds
- What happened
- Liquid AI released LFM2.5-24B-A2B, a 24-billion parameter model with only 2 billion active parameters per token, optimized for on-device agent applications.
- Why it matters
- Engineers building local AI agents or deploying reasoning models on consumer hardware like MacBooks or edge devices should evaluate this model for tool-use and agentic workloads.
- Watch out
- The model excels at tool-use benchmarks but lags on knowledge-heavy tasks like GPQA Diamond. Retrieval or web search pairing is recommended for knowledge-intensive applications.
- agent
- on-device
- lfm
Who else ran this
The same event, reported by other publishers we follow.
The patterns behind this
Each one covers how the technique works, when it earns its cost, and where it breaks.
The Agent Architect
One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.
Weekly email, one-click unsubscribe. We only use your address to send the briefing.