Glimpse launched as a research preview with API access, while its chat interface remains waitlist-only. The thesis behind the model, according to Elorian, is that today's AI systems cannot really see. Most multimodal models look at an image once, convert it to tokens and then reason about it in text, like describing a photo from memory. Glimpse instead pinches to zoom on the source mid-reasoning to double-check what it thought it saw.
CEO Andrew Dai, a former Google DeepMind researcher, described a game night as one of the clearest signs that AI still could not see. Last year he photographed a board game he had just finished with friends and asked a frontier model who was winning. It could not say. "To an adult, it's quite easy to understand what's going on with minimal text, but these models aren't able to intuit what the symbols mean," he told TechCrunch. "Humans can look at Ikea diagrams, with no text, and assemble Ikea furniture, right? Why can't AI models?"
The approach is not entirely new: OpenAI added crop-and-zoom during reasoning to its o-series models in 2025. Elorian's pitch is that it built the entire model around the idea, cheaply. The company says a small team trained Glimpse on a few hundred GPUs and that the model is substantially smaller than the frontier systems it is benchmarked against.
Glimpse costs $0.60 per million tokens in and $3 per million tokens out, cheaper than mid-level models but more expensive than lightweight ones. According to Elorian's evaluations, it outperforms GPT-6 Astra on scientific chart reasoning and short-form video captioning, with competitive results on other visual reasoning tasks. Those claims are self-reported.
Elorian came out of stealth six months ago with a round co-led by Striker Venture Partners, Menlo Ventures and Altimeter Capital, with participation from Nvidia and Google chief scientist Jeff Dean. The initial targets are companies sitting on large amounts of video: robotics firms, scientific labs, media and creative companies. Longer term, Elorian wants to move from seeing to doing, building a model of a scene and using that to choose an action, including directly controlling robots. Next year, Dai said, the company aims to build "the world's best visual reasoning model," one that leads on every visual reasoning test.