Elorian’s first AI model wants to reason with images, not just about them
Glimpse can revisit images mid-reasoning, cropping details, selecting video frames and running code to check what it sees.
Elorian’s first model is launching as a research preview with API access; its chat interface is waitlist-only. The startup says it trained Glimpse with a small team on a few hundred GPUs, charging $0.60 per million input tokens and $3 per million output tokens. In Elorian’s own evaluations, it beats GPT-6 Astra on scientific chart reasoning and short-form video captioning. Crop-and-zoom reasoning already exists in OpenAI’s o-series; Elorian’s pitch is that Glimpse was built around that approach.
TechCrunch AI's note
Elorian’s first model is launching as a research preview with API access; its chat interface is waitlist-only. The startup says it trained Glimpse with a small team on a few hundred GPUs, charging $0.60 per million input tokens and $3 per million output tokens. In Elorian’s own evaluations, it beats GPT-6 Astra on scientific chart reasoning and short-form video captioning. Crop-and-zoom reasoning already exists in OpenAI’s o-series; Elorian’s pitch is that Glimpse was built around that approach.
TechCrunch AI's note
score 6