---
id: "2026-10-16-elorian-glimpse-visual-reasoning-model"
url: "https://postcutoff.com/e/2026-10-16-elorian-glimpse-visual-reasoning-model/"
as_of: "2026-10-17T02:25:00+02:00"
date: "2026-10-16"
date_precision: day
category: model-release
importance: 2
confidence: medium
status: [Partly confirmed]
sources: 1
editor: human editor
human_review: null
version: "2026-10-16"
---

As of: 2026-10-17 02:25 CEST.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/
Canonical page: https://postcutoff.com/e/2026-10-16-elorian-glimpse-visual-reasoning-model/

# Elorian's first model, Glimpse

Full title: Elorian releases Glimpse, a visual-reasoning model that crops, re-reads and measures images while thinking, as a research preview API

Elorian, a Palo Alto startup led by former Google DeepMind researcher Andrew Dai, released its first model, Glimpse, on Oct 16, 2026 as a research preview. It uses a "visual scratch space": during reasoning it can go back to the image, crop regions, pull video frames or run code to measure. API price is $0.60/$3 per million input/output tokens; Elorian claims it beats GPT-6 Astra on scientific chart reasoning and short-video captioning (no scores published in the coverage).

## Key facts
- API available now; chat interface waitlist-only; pricing $0.60 per 1M input and $3 per 1M output tokens (TechCrunch)
- Trained by a small team on a few hundred GPUs; said to be substantially smaller than the frontier models it is compared with
- Company: out of stealth about six months earlier with a $55M seed at a $300M valuation, co-led by Striker Venture Partners, Menlo Ventures and Altimeter, with Nvidia and Jeff Dean participating
- Elorian previously co-built the Humanity's Sixth Sense visual-reasoning benchmark with Scale AI
- Target users: robotics, scientific labs, media companies with video libraries; long-term goal of scene models that drive robot control

## What happened

Glimpse is built around looking again: instead of encoding an image once, the model can re-inspect it mid-reasoning. OpenAI added a
similar crop-and-zoom ability to its o-series models in 2025; Elorian says it built the whole model around it. Benchmark claims are the
company's own and have not been independently checked; we have not found an official model page or API model id.

## Why it matters

A small, cheap visual-reasoning model from a new lab, aimed at the gap shown by its own benchmark, where people still far outscore
frontier models on intuitive visual reasoning.

## Your AI and this story

- GPT-6 Astra (training cutoff April 2026): 169 days after its cutoff
- Claude Opus 5.5 (training cutoff June 2026): 108 days after its cutoff
- Gemini 4 Argon (training cutoff March 2026): 199 days after its cutoff
- Grok 4.7 (training cutoff May 2026): 138 days after its cutoff

## Sources

1. [TechCrunch: Elorian's first AI model wants to reason with images, not just about them](https://techcrunch.com/2026/10/16/elorians-first-ai-model-wants-to-reason-with-images-not-just-about-them/) (techcrunch.com, press)

## Changes

- 2026-10-16 (filed): Created

## Related

- 2026-10-08: [Humanity's Sixth Sense (Scale AI and Elorian)](https://postcutoff.com/e/2026-10-08-humanitys-sixth-sense-visual-reasoning-benchmark/index.md)
- People: [Jeff Dean](https://postcutoff.com/person/jeff-dean/)