
Glimpse from Elorian is now available through Vercel AI Gateway. The visual reasoning model accepts images and video, follows objects through motion, interprets spatial relationships and connects actions to their physical consequences.
According to the Vercel changelog, developers can call the model using elorian/glimpse. AI Gateway provides access with one API key and no separate provider account, alongside spend tracking and request traces.
How Glimpse from Elorian handles visual evidence
Glimpse can return to the source material while it works. It can zoom into an image detail, revisit a short section of video or track the same feature across multiple frames.
That approach supports tasks where the answer depends on change over time rather than a single image. Vercel gives examples including checking whether a train is slowing down, keeping objects distinct while a camera moves and labelling completed actions in footage from a robot.
| Input or task | What Glimpse does |
|---|---|
| Images | Examines details and spatial relationships |
| Video | Revisits short intervals and follows motion |
| Moving cameras | Keeps separate objects distinct across frames |
| Robot footage | Labels actions after they have been completed |
Vercel also lists scientific chart analysis, video captioning, visual counting and robotics data annotation as supported use cases. These cover both content workflows and structured analysis, including jobs where a system needs to identify what happened and when it happened.
Using the model through AI SDK 7
Builders using AI SDK can select elorian/glimpse as the model. With AI SDK 7, video can be sent as a file part, so the model can analyse footage supplied inside an AI SDK request.
Requests made through AI Gateway include spend tracking and traces. This gives developers a record of model requests alongside the visual reasoning output, without requiring an Elorian provider account.
Frequently asked questions
What is Glimpse from Elorian on Vercel?
Glimpse is a visual reasoning model available through Vercel AI Gateway. It analyses images and video, follows objects through motion, interprets spatial relationships and connects actions with physical consequences.
How do developers call Glimpse from Elorian?
Use the model name elorian/glimpse through Vercel AI Gateway. It also works with AI SDK, including video sent as a file part with AI SDK 7.
Which Vercel plan includes Glimpse from Elorian?
The announcement does not specify plan availability or pricing. It confirms access through AI Gateway with one API key, spend tracking and request traces.
Sources
3 checkedHow we cover tool news: Create With's tool desk drafts these reports with AI from the sources listed above and checks them against those sources before publishing.








