← Catalog
LFM3-VL-2B
AvailableLiquid AIOpen weights
LFM3-VL-2B is the vision-language member of Liquid AI's LFM3 family, released 2026-10-27 alongside LFM3-1.7B. Branched from the shared LFM3-1.7B-Base tri-hybrid backbone with new multimodal encoders, it takes images and video plus text and returns text, with a 128K-token context, aimed at on-device multimodal agents. Open weights on Hugging Face under the LFM Open License v1.0; also on the Liquid Playground/API. Exact parameter count not recorded (2B class).
Specifications
- License
- Open weights · LFM Open License v1.0
- Weights
- Downloadable
- Architecture
- hybrid
- Parameters
- Undisclosed
- Context window
- 131K tokens
- Max output
- —
- Knowledge cutoff
- —
- Price (in / out, $/M)
- —
- Modalities
- TextVisionVideo
Benchmarks
No benchmark scores recorded yet. Spotted some? Submit a correction.
Vendor-reported figures are claims until independently verified. See methodology.