LFM3-1.7B is the text model in Liquid AI's LFM3 generation (1.7B parameters, 256K-token context, LFM Open License), launched October 2026 on a tri-hybrid architecture of short convolutions plus sliding-window and global attention. Post-trained for agents, tool calling, and structured outputs, it reasons before answering and fits in about 1 GB at 4-bit for on-device use, though Liquid does not recommend it for agentic coding or complex multi-step reasoning.
コスト率
0.1x
コンテキスト
262K
リリース
2026年11月6日
入力
Text
出力
Text
対応
推論ツール呼び出し構造化出力