{"apiVersion":"v1","taxonomyVersion":1,"generatedAt":"2026-10-29T15:43:34.124Z","data":{"recordId":"wsa_yyhrd9cyf48a","score":1,"confidence":"high","capped":false,"capReason":null,"sources":[{"n":1,"url":"https://huggingface.co/LiquidAI/LFM2.5-24B-A2B","publisher":"Hugging Face","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-27T01:17:35.750Z","detectedPublishedAt":"2026-10-12T05:35:46.000Z","rights":"original"},{"n":2,"url":"https://huggingface.co/LiquidAI/LFM2.5-24B-A2B-MLX-4bit","publisher":"Liquid AI","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-27T01:17:35.750Z","detectedPublishedAt":"2026-10-12T19:31:28.000Z","rights":"original"},{"n":3,"url":"https://www.liquid.ai/blog/lfm2-5-24b-a2b","publisher":"Liquid AI","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-27T01:17:35.750Z","detectedPublishedAt":"2026-10-13T00:00:00.000Z","rights":"original"},{"n":4,"url":"https://huggingface.co/blog/LiquidAI/lfm25-24b-a2b","publisher":"Hugging Face","authority":"other","verification":"verified","lastRetrievedAt":"2026-10-13T22:16:11.424Z","detectedPublishedAt":"2026-10-13T17:05:48.000Z","rights":"original"}],"claims":[{"id":"cb75464e-9eb3-4c3b-854d-f29a7edf6998","kind":"deployment","statement":"The model was converted to MLX format using mlx-lm version 0.32.0.","citations":[{"source":2,"supportProbability":0.98,"supports":true,"quote":{"text":"This model [LiquidAI/LFM2.5-24B-A2B-MLX-4bit](https://huggingface.co/LiquidAI/LFM2.5-24B-A2B-MLX-4bit) was converted to MLX format from [LiquidAI/LFM2.5-24B-A2B](https://huggingface.co/LiquidAI/LFM2.5-24B-A2B) using mlx-lm version **0.32.0**.","rights":"original"}}]},{"id":"fe1fed1d-53ac-4447-98e2-2df77f8d5b26","kind":"capability","statement":"The model is quantized to 4 bits per weight, with token embeddings kept at 6 bits and MoE routing gates kept unquantized in float32.","citations":[{"source":2,"supportProbability":0.97,"supports":true,"quote":null}]},{"id":"7b19da40-efec-44ec-bfd2-5d5e12c574af","kind":"availability","statement":"Loading the model requires mlx-lm 0.32.0 or later.","citations":[{"source":2,"supportProbability":0.82,"supports":true,"quote":{"text":"Requires mlx-lm 0.32.0 or later.","rights":"original"}}]},{"id":"244486e9-581f-4fd8-ab31-a8ffcefeece5","kind":"recommendation-relevant","statement":"Earlier mlx-lm versions load the model but route the experts incorrectly.","citations":[{"source":2,"supportProbability":0.97,"supports":true,"quote":{"text":"Earlier versions load the model but route the experts incorrectly.","rights":"original"}}]},{"id":"f9f704b3-7932-40ac-a31f-3aa8cb894696","kind":"capability","statement":"The model has 24B total parameters and 2.45B active parameters.","citations":[{"source":1,"supportProbability":0.98,"supports":true,"quote":{"text":"- **Best-in-class agent**: Competitive with Qwen3.6-35B-A3B, a model with ~50% more total and active parameters, on agentic and tool-use tasks, scoring 48.09 on LiveClaw versus 38.19.","rights":"original"}}]},{"id":"0a920edf-d6a9-47b4-8491-9ce19f89b06b","kind":"capability","statement":"The model has a context length of 131,072 tokens.","citations":[{"source":1,"supportProbability":0.99,"supports":true,"quote":{"text":"- **Context length**: 131,072 tokens","rights":"original"}},{"source":3,"supportProbability":0.96,"supports":true,"quote":null}]},{"id":"c76d0838-52cd-4e90-961a-513a5528ee31","kind":"capability","statement":"LFM2.5-24B-A2B supports function calling for agentic workflows.","citations":[{"source":1,"supportProbability":0.99,"supports":true,"quote":{"text":"LFM2.5-24B-A2B supports tool calling for agentic workflows.","rights":"original"}}]},{"id":"a86bcd8b-3655-42f7-82aa-291537dc7513","kind":"availability","statement":"The original model checkpoint is available in native format on Hugging Face.","citations":[{"source":1,"supportProbability":0.95,"supports":true,"quote":{"text":"| **[LFM2.5-24B-A2B](https://huggingface.co/LiquidAI/LFM2.5-24B-A2B)**               | Original model checkpoint in native format.","rights":"original"}}]},{"id":"1a49d95f-dd1b-4c1b-adc9-97a10e11c7fe","kind":"deployment","statement":"The model can be run locally with inference frameworks including Transformers, vLLM, SGLang, and llama.cpp.","citations":[{"source":1,"supportProbability":0.84,"supports":true,"quote":null}]},{"id":"5075737e-fc91-4dba-87c9-6e6a26a6e5ee","kind":"recommendation-relevant","statement":"The source recommends pairing the model with retrieval or web search for knowledge-heavy tasks.","citations":[{"source":1,"supportProbability":0.97,"supports":true,"quote":{"text":"It is not recommended for knowledge-heavy tasks on its own: pair it with retrieval or web search, which plays to its tool-calling strength.","rights":"original"}},{"source":3,"supportProbability":0.98,"supports":true,"quote":null}]}]},"nextCursor":null,"license":{"catalog":{"name":"CC BY 4.0","spdx":"CC-BY-4.0","url":"https://creativecommons.org/licenses/by/4.0/","attribution":"wowsuchai","attributionUrl":"https://wowsuchai.withglint.app","covers":["catalog metadata","classifications","generated summaries","feed event metadata"]},"excluded":{"covers":["source quotations","linked source content"],"terms":"Not licensed by wowsuchai. Source quotations and linked source content keep their original rights."}}}