Elorian’s first AI model wants to reason with images, not just about them
Elorian released Glimpse, a research-preview model that revisits images mid-reasoning through a visual scratch space.
Palo Alto startup Elorian released Glimpse, its first model, as a research preview with API access and a waitlisted chat interface. Glimpse is built around a “visual scratch space” that can crop image regions, pull video frames, or run measurement code during reasoning rather than converting an image to tokens once. Elorian says a small team trained the comparatively small model on a few hundred GPUs. It prices Glimpse at $0.60 per million input tokens and $3 per million output tokens, and claims it outperforms GPT-6 Astra on scientific chart reasoning and short-form video captioning.
58