LiquidAI has released LFM2.5-24B-A2B, a new reasoning-focused language model designed for efficient on-device agent deployment. Despite having fewer parameters than competitors like Qwen3.6-35B-A3B, it demonstrates competitive performance on agentic tasks and tool use, achieving high inference speeds on both CPUs and GPUs. The model's architecture builds upon its predecessor with an expanded context window and vocabulary, and it undergoes a multi-stage post-training process to enhance its agentic capabilities. AI
IMPACT Enables faster, more private AI agents on consumer hardware by optimizing for efficient inference.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- Apple M5 Max
- AutomationBench
- Gemma-4-26B-A4B-it
- Hermes Agent
- LFM2-24B-A2B
- LFM2.5-24B-A2B
- LiquidAI
- LiveClaw
- NVIDIA B300 GPU
- OpenCode
- Qwen3.6-35B-A3B
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
