Elorian makes Glimpse available for visual reasoning through the API
Elorian has launched Glimpse as a research preview with API access. During reasoning, it can examine image crops, select video frames and run code. For now, the chat interface only has a waitlist.
Elorian has released Glimpse, its first model for visual reasoning. It is available as a research preview through the API; only a waitlist is open for the chat interface. Pricing is 0.60 USD per million input tokens and 3 USD per million output tokens.
During reasoning, the model can repeatedly revisit the image material: cropping regions of an image, selecting specific video frames or running code to measure and compare. Elorian calls this mechanism “visual scratch space”.
According to an evaluation by Elorian, Glimpse outperforms GPT-6 Astra in reasoning about scientific charts and generating captions for short videos. This comparison therefore comes from the manufacturer. Elorian is initially targeting robotics companies, scientific laboratories, and media or creative companies that work with large amounts of video.
Why it matters
API access makes it possible to try image and video analysis in which the model checks details in the original material throughout the process. The practical benefit for a specific task needs to be verified using your own data, because the published performance results come from an evaluation by Elorian.
Release card
Glimpse
Elorian
- Inputs
- The model works with images and video and generates text captions for videos.
- Price
- 0.60 USD per million input tokens and 3 USD per million output tokens
- Availability
- The model is available through the API as a research preview. For the chat interface, only a waitlist is open for now.
- Reasoning about scientific charts.
- Generating text captions for short videos.
- Measuring and comparing image material using code.
According to an evaluation conducted by Elorian itself, Glimpse outperforms GPT-6 Astra in reasoning about scientific charts and generating captions for short videos. Elorian emphasizes that the model repeatedly revisits the image material during reasoning.
The card summarizes information from the article and any dated corrections, with a link to the original source. It is not our assessment of the model. It does not yet have a dedicated editorial profile. Model selection and other announcements →
Two audiences, two different impacts
What this means
For individuals
For individuals who want to work through the chat interface, the model launch does not yet mean immediate access: only a waitlist is available.
For a business
Teams processing video or scientific charts can integrate the model into a trial workflow through the API and assess both output quality and costs based on the volume of input and output tokens.
ProcessesCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.