Elorian Ships First Model Glimpse, Claiming Vision Edge

Elorian, a Palo Alto AI startup that raised a $55 million seed round at a $300 million valuation before shipping anything, released its first model on Friday, TechCrunch has exclusively learned.

The model, called Glimpse, is a research preview. API access is open, but the chat interface is waitlist-only. So far, the evidence amounts to a results table and a thesis.

What does Glimpse actually do differently?

The thesis is blunt: today's AI models, for all their skill with language and code, still 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 keeps going back to the source. Mid-reasoning, it can crop into a region, pull specific frames from a video, or run code to measure and compare what it is looking at. Elorian calls this mechanism a "visual scratch space." The company likens it to someone pinching to zoom on a phone photo to double-check what they thought they saw.

The takeaway: Elorian bets that vision is an architecture problem, not a scale problem.

Why the CEO thinks rivals can't see

For former Google DeepMind researcher and Elorian CEO Andrew Dai, the clearest sign that AI still could not really see came on game night. 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.

Board games "each have their own visual language, each have their own iconography, symbols," Dai told TechCrunch. "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."

"Humans can look at IKEA diagrams, with no text, and assemble IKEA furniture, right? Why can't AI models?" he said.

The takeaway: the anecdotes are vivid, but they illustrate a failure mode, not a benchmark.

How cheap is it, and how small?

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 whole model around that mechanism, cheaply. The company says a small team trained Glimpse on a few hundred GPUs, and the model is substantially smaller than the frontier systems it benchmarks against.

Pricing sits in the middle of the market:

• $0.60 per million input tokens
• $3 per million output tokens
• Cheaper than mid-level models like Sonnet and Sol
• More expensive than lightweight models like Haiku and Luna

The takeaway: Glimpse undercuts the mid-tier it claims to beat, which will pressure rivals if the benchmarks hold.

Whose numbers are these?

According to Elorian's own evaluations, Glimpse outperforms even GPT-6 Astra on scientific chart reasoning and short-form video captioning, with competitive results on other visual reasoning tasks. Those claims come from the company itself. No independent verification accompanies the launch, and the waitlist-only chat interface limits outside scrutiny for now.

The takeaway: self-reported wins on selected tasks warrant caution until third parties test the model.

Who is behind Elorian, and what comes next?

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 it 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. That promise covers a lot of ground, and a small team on a few hundred GPUs has to beat labs with vastly larger compute budgets to keep it.

By Olivia Hart