d1-Omni-2B
d1-Omni-2B adds audio input to Liquid's 2B decision line.
What the release provides
The 2.17B checkpoint builds on LFM3-Omni-2B and handles typed decisions over text, pictures, video and audio. It returns answers in one pass without generating tokens. Its card gives a 128K context and LFM Open License terms.d1-Omni-2B official model card
How to use it well
Practical guidance. Compare spoken and written versions of the same request. Keep recognition errors separate from incorrect selection among the allowed outcomes.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | LiquidAI/d1-Omni-2B |