LFM2 1.2B
Very small. Runs on the laptop you were about to recycle.
Model / Liquid AI
Trust score 64 of 100. No, open weights with custom rules.
Last checked Changed Oct 15, 2026 History
Short answer (as of Oct 15, 2026): LFM2.5 24B-A2B is a local AI model from Liquid AI. Truly open? No, open weights with custom rules. Business use: yes, with limits. Our estimate says it needs about 20 GB of RAM. Trust score 64 of 100.
As of Oct 15, 2026: license LFM Open License v1.0; business use yes, with limits; sign-up not needed; about 20 GB RAM (estimate); trust score 64/100. Sources.
Fast local agents that think before they answer. Built for laptops.
Who makes it: Liquid AI. The files come from the maker's own Hugging Face account.
License: LFM Open License v1.0. Liquid AI's own license. It is free for business use only while your company makes under 10 million US dollars a year.
Business use: Only if your company makes under 10 million US dollars a year. Bigger companies need a deal with Liquid AI.
Training data: We have not found a full public data release. Not yet checked in depth.
Pick a computer with at least 32 GB of memory. On a Mac, that is the unified memory. On a PC, a graphics card with that much VRAM is fastest. A PC with only normal RAM will also work, just slower. This model uses a mix of experts. Only part of it works on each word, so it runs faster than its size suggests.
LiquidAI/LFM2.5-24B-A2B-GGUF.Q4_K_M. It is the best mix of size and quality for most people.The files are GGUF builds that LM Studio can load. We checked that this repo exists on Oct 15, 2026. Full LM Studio guide.
Does it fit? Yes. Its memory window is big enough for Hermes.
hermes model. Choose LM Studio and press Enter to keep http://localhost:1234/v1.hermes to start.No, open weights with custom rules. Liquid AI's own license. It is free for business use only while your company makes under 10 million US dollars a year.
Only if your company makes under 10 million US dollars a year. Bigger companies need a deal with Liquid AI.
No. Download without an account.
Our estimate is about 20 GB of RAM for a 4-bit file of about 14.4 GB. This is a rough guide, not a test.
| Fact | What we found | Status |
|---|---|---|
| Maker | Liquid AI | Verified Oct 15, 2026 |
| License | LFM Open License v1.0 | Verified Oct 15, 2026 |
| License plain | Liquid AI's own license. It is free for business use only while your company makes under 10 million US dollars a year. | Verified Oct 15, 2026 |
| Truly open | No, open weights with custom rules | Verified Oct 15, 2026 |
| Business use | Only if your company makes under 10 million US dollars a year. Bigger companies need a deal with Liquid AI. | Verified Oct 15, 2026 |
| Signup needed | No. Download without an account. | Verified Oct 15, 2026 |
| Parameters | 24.0 billion | Verified Oct 15, 2026 |
| Open training data | Not yet checked | We have not found a full public training data release. Not yet checked in depth. |
| Context window | 128K tokens | Verified Oct 15, 2026 |
| First posted | 2026-10-12 | Verified Oct 15, 2026 |
| Gguf files | LiquidAI/LFM2.5-24B-A2B-GGUF | Verified Oct 15, 2026 |
| Hermes ok | Yes | Verified Oct 15, 2026 |
If the maker's page and this page disagree, trust the maker. Tell us what we got wrong.
Newsletter
Which new models are worth your time, which ones are just hype, and the license gotchas before they bite. All the useful bits, none of the nerd-party small talk.
This opens our sign-up page on Beehiiv, our email provider. Confirm your email there to join.