d1-omni-600M
Released Oct 7, 2026 · knowledge cutoff unpublished
- Status
- Unknown
- Location
- United States
- Type
- System One
- Modality
- Multimodal
- Context window
- 16K
- Max output
- —
- Speed
- —
- Price ($/MTok in / out)
- — / —
- Cost per task
- —
- Open weights
- Yes
Strengths & weaknesses
Strengths
- Takes audio as well as images alongside text, so spoken input can be classified without a separate transcription step
- Small enough for phones and embedded boards where the 3B sibling won't fit — community CoreML and LiteRT ports appeared within days
- Same one-pass, zero-output-token design as d1-3B, returning calibrated probabilities for yes/no, choice and score questions
Weaknesses
- Liquid calls it an experimental checkpoint, an early research release rather than a production model
- Takes text with images or text with audio, not all three together, and image inputs cut the state and question text to 896 tokens as trained
- Audio decisions have no benchmark yet — Liquid calls audio decision evaluation an open problem and publishes no speed figures for this model
Not an LLM: the smaller of Liquid AI's two open "System One" decision models released on Hugging Face on 2026-10-07, 587M parameters built on LFM2.5-Encoder-350M, returning calibrated probabilities for typed questions in one forward pass, so none of this site's benchmarks apply and every cell is null by design. Context is 16,384 tokens covering text, image and audio positions per the model card (its 65,536 figure is vocabulary size, not context). Same LFM Open License v1.0 terms as d1-3B; Liquid's seven-dataset text mean of 78.4 is its own measurement. Availability is self-host: weights only, with community GGUF, CoreML and LiteRT conversions but no hosted endpoint. Sources are search excerpts of liquid.ai/blog/d1-open, huggingface.co/LiquidAI/d1-omni-600M and docs.liquid.ai, not pages read directly.
Benchmarks0 of 10 reported
MMLU-Proretired
GPQA Diamond
SWE-bench Verified
SWE-bench Pro
Terminal-Bench 2.1
AIME (latest)retired
Humanity's Last Exam
LMArena Elo
GDPval-AA v2
ARC-AGI-2
For developers
API model strings
Not researched
Licence
LFM Open License v1.0 · Restricted · commercial use permitted
Retirement
No retirement announced
Lineage
No recorded predecessor