Side-by-side comparison of LFM2.5-8B-A1B (Liquid AI) and Llama 3.1 Nemotron 70B (NVIDIA · USA) for self-hosted deployment of the open-weight model. LFM2.5-8B-A1B is rated conditional; Llama 3.1 Nemotron 70B is conditional. They part ways on licence: LFM2.5-8B-A1B is "LFM Open License 1.0", Llama 3.1 Nemotron 70B is "Llama community".
| Field | ||
|---|---|---|
| Summary | ||
| Verdict | Conditional Per the published LFM Open License v1.0, Liquid AI's terms are Apache-2.0-derived but cap free commercial use at USD 10M annual company revenue — above that threshold deployers must negotiate a paid licence with [email protected]. The model itself is a strong 8.3B / 1.5B-active MoE built for CPU and edge inference, but the 38-trillion-token training corpus is undisclosed and Liquid AI is a US-based MIT spinoff, so EU regulated buyers should run their own DPIA and document the data-transfer story alongside any commercial-tier negotiation. | Conditional NVIDIA's Llama 3.1 fine-tune with custom RLHF. Inherits Llama 3.1 Community License terms. Strong conversational quality; useful default when you want Llama behaviour with NVIDIA's alignment. |
| Last reviewed | 2026-05-30 | 2026-04-15 |
| Open-weight | ||
| Licence | LFM Open License 1.0 | Llama community |
| Commercial use | Free under $10M revenue | With caps |
| Training data | Undisclosed | Partial |
| Origin | USA (Boston) | USA |
| Performance & pricing? | ||
| Quality index | — | 13/100 |
| Speed | — | 42 tok/s |
| Blended price | — | $1.20/M |
| Context window | — | — |
| Evidence | ||
| Sources | ||
No overlapping sources between the two entries.