Model release

LFM2.5-24B-A2B

LFM2.5-24B-A2B is an open-weight model release in the LFM (Liquid Foundation Models) family from Liquid AI1 4, version LFM2.5-24B-A2B, released Oct 13, 2026. Its weights can be downloaded by the public1 4 under the LFM Open License v1.02 1.

A text-only reasoning mixture-of-experts model with 24.0B total and 2.45B active parameters, built from 30 double-gated short-convolution blocks and 10 grouped-query attention blocks, with 64 experts per MoE block and top-4 routing. It keeps the LFM2-24B-A2B architecture but extends the context window to 131,072 tokens and doubles the vocabulary to 128,000 tokens; unlike its predecessor, it always writes a chain of thought before answering. Liquid AI describes it as the final model in the LFM2.5 family.1 4

Model hub: Model card (Hugging Face) (external site: huggingface.co)
License: License (LFM Open License v1.0) (external site: huggingface.co)
Release notes: Release post (external site: liquid.ai)

On AI on Hugging Face: LiquidAI/LFM2.5-24B-A2B

Last reviewedOct 18, 2026Entry updated Oct 18, 2026Documented release Oct 13, 2026

Availability and license

Overall availability

Public

Weights download from Hugging Face without an access gate; use is governed by the LFM Open License v1.0, which conditions commercial-use rights on the licensee's legal entity not exceeding a threshold defined as annual revenue of US$10,000,000 or more. Liquid also offers the model in its Playground and through its paid Liquid Console API.1 2 4

Availability fact review: not separately recorded

Availability is separate from permission: read the license before using or redistributing.

LFM Open License v1.0 (external site: huggingface.co)2 1
Applies to: Weights · custom license (no SPDX identifier) · fact-level rights review not recorded
Plain-language guide to the LFM Open License v1.0

Custom Liquid AI license (card license field "lfm1.0") modeled on Apache 2.0. It grants perpetual, royalty-free copyright and patent licenses, requires redistributors to include the license and mark modified files, and ends patent rights for anyone who files patent litigation over the work. Section 5 makes commercial rights conditional on the licensee's legal entity not exceeding a "Threshold" defined as annual revenue of US$10,000,000 or more, and states that commercial use by a legal entity that exceeds the Threshold is not licensed. The Threshold does not apply to a qualified non-profit organization's use for non-commercial or research purposes. Liquid's license page FAQ says fine-tuned models may be kept proprietary and that hosting platforms may host and distribute the models.2 3

Component reuse rights

A readable or downloadable component is not automatically reusable. These indicators concern recorded license evidence, not system certification.

weights: Unknown — no complete fact-level rights review
code: Unknown — no complete fact-level rights review
data: Unknown — no complete fact-level rights review
documentation: Unknown — no complete fact-level rights review

No complete system-rights review is recorded for this release.

Model-disclosure tier

Computed from the checklist below using USASI rubric v0.2. An editorial category, not a certification.

Open-weight

The model parameters for this release can be downloaded by the public. License terms may still restrict use, redistribution, or commercial use.

The weights are under a license that is not on the rubric's OSI-approved list. Read its terms before use.

How tiers are computed

Public materials checklist

Items for a model under USASI rubric v0.2. Unknown means unassessed or insufficient evidence.

Public materials checklist for LFM2.5-24B-A2B
Item	Status	Notes and evidence
WeightsCan the general public download the model parameters for this release?	Public	Ungated safetensors weights on Hugging Face; a base checkpoint and FP8, GGUF, ONNX, and MLX 4-bit versions are published as separate repositories.1 4
Inference codeIs code for running the model published?	Public	The model card documents inference with Hugging Face Transformers (5.0.0 or later), vLLM, SGLang, llama.cpp, MLX, and LM Studio, and gives setup examples for several agent harnesses; no custom modeling code is needed in the repository.1
Training codeIs the code used to train the model published?	Unknown	No training code release was found in the model card or release post. Liquid documents fine-tuning with third-party tools (Unsloth, TRL), which is not training code for this release.
Training-data informationDoes the information cover provenance, scope, acquisition, selection, labeling, processing, and where data or alternatives can be obtained? Access alone does not establish completeness.	Unknown	Not assessed.
Training-data accessCan the training data be obtained? This is independent of information completeness and reuse rights; original unshareable data need not be downloadable.	Unknown	Not assessed.
Complete training pipelineIs the complete base-training and preprocessing pipeline published, including configuration? Fine-tuning code or an inference SDK alone is insufficient.	Unknown	Not assessed.
Legacy data assessment (v0.1)Historical assessment combining download access and disclosure. Preserved for traceability; excluded from the v0.2 tier calculation. See the new separate assessments above.	Unknown	The model card and release post give token budgets (about 33T pretraining tokens plus mid-training phases) and broad data emphases, but do not describe data sources or composition.
Training recipeAre the training configuration and procedure documented in enough detail to follow?	Partial	The model card and release post describe the stages at a high level: pretraining, two mid-training phases (2T tokens at 32K context, then 300B tokens at 131K context), and a five-stage post-training pipeline of supervised fine-tuning, teacher specialization, two on-policy distillation stages, and agentic reinforcement learning inside agent harnesses. Hyperparameters and data mixes are not given.1 4
Evaluation materialsPublic = evaluation code or prompts that let others re-run the evaluations are published. Partial = results only.	Partial	The model card and release post report benchmark tables and the release post lists generation settings used; evaluation code or prompts are not linked.1 4

What it is useful for

The model card recommends it for on-device agents that reason and call tools, and says it is not recommended on its own for knowledge-heavy tasks, suggesting it be paired with retrieval or web search instead.1 4

Run and use notes

Documented facts only. No hardware or performance claims are made without a cited source and stated assumptions.

The model card recommends temperature 0.1, top_k 50, and repetition_penalty 1.05, and shows Transformers loading in bfloat16. It lists a DSpark speculative-decoding drafter for SGLang as coming soon.1

Organization context

Liquid AI

Provenance and derivatives

Post-trained by Liquid AI from its own LFM2.5-24B-A2B-Base checkpoint, which it pretrained itself and released alongside this model; no third-party base model is involved.1 4

Derived from: LFM2.5-24B-A2B-Base (external site: huggingface.co) — Same organization; pretrained base checkpoint.

Other releases in the LFM (Liquid Foundation Models) family

LFM2.5-2.6BModel-disclosure tier (USASI rubric v0.2): Open-weight
LFM2.5-8B-A1BModel-disclosure tier (USASI rubric v0.2): Open-weight

LFM (Liquid Foundation Models) family overview

What this catalog does not know

Unknown means the sources reviewed for this record do not document it. It is not evidence that something does not exist.

Training code: unknown.
Training-data information: unknown.
Training-data access: unknown.
Complete training pipeline: unknown.
Legacy data assessment (v0.1): unknown.

Have a primary source? How to report a correction.

U.S. eligibility

Project eligibility rests on documented governing or maintaining entities, not on contributors.

Eligible · basis: U.S. headquarters

Developed and published by Liquid AI, Inc., the licensor named in the model's license file. Liquid AI gives its address as Cambridge, Massachusetts (see the liquid-ai record).1 2 5

Assessed Oct 18, 2026

Sources

1.
LiquidAI/LFM2.5-24B-A2B model card (external site: huggingface.co)
Liquid AI · Model card · accessed Oct 18, 2026
2.
LFM2.5-24B-A2B LICENSE (LFM Open License v1.0) (external site: huggingface.co)
Liquid AI · License · accessed Oct 18, 2026
3.
LFM License (external site: liquid.ai)
Liquid AI · License · accessed Oct 18, 2026
4.
LFM2.5-24B-A2B: Built for Fast On-Device Agents (external site: liquid.ai)
Liquid AI · Announcement · published Oct 13, 2026 · accessed Oct 18, 2026
5.
Liquid AI Privacy Policy (external site: liquid.ai)
Liquid AI · Official page · published Sep 30, 2026 · accessed Oct 18, 2026

Email a correction Report a correction Edit this entry

This listing is not an endorsement, a safety assessment, or a federal approval.

Cite this page

APA style

USASI. (2026, October 18). LFM2.5-24B-A2B. United States of America Superintelligence. https://unitedstatesofamericasuperintelligence.com/open/lfm2-5-24b-a2b/

BibTeX

@misc{usasi_lfm2_5_24b_a2b,
  author = {{USASI}},
  title = {LFM2.5-24B-A2B},
  year = {2026},
  month = oct,
  howpublished = {\url{https://unitedstatesofamericasuperintelligence.com/open/lfm2-5-24b-a2b/}},
  note = {United States of America Superintelligence. Last updated 2026-10-18}
}

The date is when this page was last updated. Add the date you read it if your style needs one.

For maintainers: a “Listed in USASI” badge

Links to this page. It says the entry is listed, not endorsed or rated. About the badge

Markdown (for a README)

[![Listed in the USASI catalog](https://unitedstatesofamericasuperintelligence.com/badges/listed.svg)](https://unitedstatesofamericasuperintelligence.com/open/lfm2-5-24b-a2b/)

HTML

<a href="https://unitedstatesofamericasuperintelligence.com/open/lfm2-5-24b-a2b/"><img src="https://unitedstatesofamericasuperintelligence.com/badges/listed.svg" alt="Listed in the USASI catalog" height="20"></a>