New model: Glimpse Maker: Elorian It's a "visual thinking" model that reasons about objects, motion and viewpoints from images and video. Elorian claims multimodal scores on par with GPT-6 Astra and Claude Opus 5.5 from a much smaller model, post-trained on an open-weights base. Detected via Vercel AI Gateway.
Model reference spotted ·
elorian/glimpse
by Elorian found in Vercel AI Gateway model catalog
Model reference
- Status
- Listed in a catalog, not announced
- Maker
- Elorian
- Found in
- Vercel AI Gateway model catalog
- Detected
- Posted to X
- 0 s after detection
- Model ID
elorian/glimpse
Where to use it
first seen Oct 16, 2026| Where | Listed | After first sighting |
|---|---|---|
| OpenRouter $0.60 in / $3.00 out per 1M tokens · 262K context | +3 h |