models · liquidai
text-generation model harvested from Hugging Face (trending). Catalogued at Quarantine — unverified, pending attestation.
| Method | hf-published-sha256 |
|---|
| Publisher | liquidai |
|---|---|
| Origin | United States |
| Licence | other |
| Parameters | 8,467,856,832 (8.47B) |
| Architecture | lfm2_moe |
| Layers | 24 |
| Hidden size | 2048 |
| Attention | GQA (32 heads, 8 KV) |
| Native precision | BF16 |
| Pulls recorded here | 142,145 |
| Downloads reported upstream | 80,632 |
| GPQA Diamond | 51.3 |
|---|---|
| IFBench | 55.6 |
| Humanity's Last Exam | 6.9 |
Reported by a third party and recorded here, not re-run by us. Source: artificialanalysis.ai/api/v2.
| Hardware | Build measured | Decode @ 8K | Memory @ 8K | Decode mode |
|---|---|---|---|---|
| NVIDIA GeForce RTX 5090 32GB | Q8_0 | 509 tok/s | 13.2 GB | ordinary |
| NVIDIA RTX PRO 6000 Blackwell 96GB | Q8_0 | 485 tok/s | 13.3 GB | ordinary |
| NVIDIA GeForce RTX 5090 32GB | — | 399 tok/s | 30.3 GB | ordinary |
| NVIDIA GeForce RTX 5090 32GB | default | 399 tok/s | 30.3 GB | ordinary |
| NVIDIA GeForce RTX 5090 32GB | unknown | 399 tok/s | 30.3 GB | ordinary |
| NVIDIA RTX PRO 6000 Blackwell 96GB | — | 390 tok/s | 90.6 GB | ordinary |
| NVIDIA RTX PRO 6000 Blackwell 96GB | default | 390 tok/s | 90.6 GB | ordinary |
| NVIDIA RTX PRO 6000 Blackwell 96GB | unknown | 390 tok/s | 90.6 GB | ordinary |
Measured by local.ai, not by us: 46 runs, 38 more not listed. Each names its engine pinned by image digest and the exact serve command, in the full record.
Registry record as of 2026-09-01. Identity, licence and parameter count are properties of a release and do not move; tier can, and the verify link above re-checks it against the live registry rather than this page.