Microsoft AI's real-time speech-to-text model for voice agents, transcribing 60 languages with continuous language detection
Glimpse
Elorian's visual thinking model, a research preview released 16 October 2026 that reasons over images and video with tools and subagents, at $0.60 in and $3.00 out per million tokens.
TrustList summary
Not yet independently verified. Every detail here comes from Elorian's own announcement of 16 October 2026; we have not yet read an independent report. The benchmark comparisons are Elorian's own measurements, so no score is stored. We will update this when it can be confirmed, and remove this note.
Glimpse is a multimodal model from Elorian, released as a research preview on 16 October 2026 and available the same day through Elorian's API and its chat interface. Elorian calls it a visual thinking model: one that reasons about the physical world from what it sees, rather than turning an image into text and reasoning over the text.
What it does
Glimpse follows objects in motion and across changing viewpoints, works out how parts fit together, and connects an observation to what is likely to happen next. Elorian says it can revisit and refine internal visual representations while it thinks, so it can zoom into a region, watch a short stretch of video at a denser frame rate, or track one feature over time before answering, and the chat interface shows that visual scratch space. As an agent, it is trained end to end to use multimodal tools that write code, transform images and watch videos, and it can start and coordinate subagents that exchange text, visual artefacts and internal representations.
Who it is for
Elorian points to robotics, scientific work and the creative industries, with examples ranging from analysing laboratory charts and tagging short social videos to identifying completed actions in robot footage. Its announcement compares Glimpse with larger frontier models on manipulation-event understanding, embodied spatial reasoning, chart reasoning, video captioning, temporal reasoning and counting, claiming it leads on four of those tests and matches on embodied reasoning; those are Elorian's own results under its own harness, and none is stored as a score here. Elorian says the model was built by a small team on a few hundred GPUs and post-trained on an open-weights base model, and that token efficiency is its next priority because the model reads inputs closely and spends tokens doing so.
Key facts
Maker: Elorian Released: 16 October 2026 (research preview) Inputs: images, video and text Access: Elorian API and chat UI Price: $0.60 per million input tokens (uncached), $3.00 per million output tokens Weights: not published; post-trained on an open-weights base model
Benchmarks & AI stats
| Benchmark | Official | Community avg |
|---|---|---|
| No benchmark scores yet. Be the first to add one. | ||
“Official” values are editor-approved and feed the ranking. “Community avg” is the mean of member submissions (shown for transparency; it never affects the ranking until an editor approves a value).
Sign in to add a benchmark score for this model.
Glimpse launches
Request a demo or quote from Glimpse
Reviews
Questions & answers
No questions yet. Be the first to ask about Glimpse.
Glimpse alternatives
Similar listings buyers compare against this one.
Ai2's open-weight 8B model that writes cited scientific reports from a research question and retrieved literature
Black Forest Labs' image generation and editing model with bounding-box layouts, multiple reference images and 4K output
Open-source 2B decision model from AWS's Strands Agents project that picks options and returns scores with a reliability figure
Cantina Security's first open-weight security model, a post-trained GLM-5.3-Flash for vulnerability research and exploit checks
Anthropic's smallest and fastest Claude 5.5 model, available on its own API and three cloud platforms
Aleph Alpha's open-weight English-German mixture-of-experts model for self-hosting, released under Apache 2.0 in October 2026
JetBrains' open 12B mixture-of-experts coding model with 2.5B parameters active per token under Apache 2.0
More on TrustList
Everything here links back to the same verified catalogue. Pick your next stop.
- Write a reviewHelp other buyers decide
- CompaniesAgencies, consultancies and IT service providers, ranked by verified reviews.
- ProductsSoftware and SaaS with pricing, features, integrations and alternatives.
- ArticlesGuides, comparisons and research from the editorial desk and the community.
- AwardsAnnual recognition decided by verified reviews and an independent jury.
- LaunchesNew products and releases, voted up by the community every day.
- RequestsBuyers describe what they need; vendors respond directly.
- PeopleReviewers, authors and makers with public profiles.
- ComparePut up to four listings side by side before you shortlist.