V
Visko
Live Models to predict the world
About Visko
Visko is an AI research lab building persistent “live models” that stream interactive visual worlds continuously instead of generating short, offline video clips. Founded by CEO Qing Yin and advised by UC Berkeley professor Michael I. Jordan, its first foundation model Orbis runs an internal world clock in a unified latent space of perception, memory and physics tokens governed by neural stochastic differential equations, so users can add or change prompts mid-stream while the model keeps visual and physical coherence for hours at 4K and 24 fps—addressing temporal degradation that breaks conventional query-based video generators. Orbis is open for public play on Visko’s site and, in company evaluations on DOVER, VideoAlign and a human preference Arena, ranked ahead of other real-time and long-video systems including Odyssey-2. In September 2026 Visko raised $10 million in pre-seed funding from Llama Capital to advance live-model research and productise Orbis for applications from creative media and gaming to robotics and physical simulation. Sources: SiliconANGLE 1 Sep 2026; visko.ai.
Source: Dealroom company dataAt a glance
- Status
- Operational
- Headquarters
- San Francisco, United States
- Employees
- 6
- Open positions
- None listed
- Dealroom signal
- 0.0/100
- Industry
- Media, Enterprise Software
- Sub-industry
- Content production
- Sector
- Other B2B Software, Generative AI, genAI Model Maker, genAI Applications
- Business model
- Manufacturing
- Client focus
- b2c
Knowledge graph
More headlines
Funding history
Company financing events and reported valuations.
| Deal type | Date | Amount | Valuation | Investors | Source |
|---|---|---|---|---|---|
| SEED✓Equity / VC | Sep 2026 | $10M | $40M | Llama Venture | ↗ |
Similar companies
Company identity and context are open; selected comparison metrics preview what is available on the platform.
| Company | Headquarters | Status | Employees | Total funding | Valuation |
|---|---|---|---|---|---|
 Beijing Aishi Technology | PixVerseGenerative AI for multimodal video creation |
China | Operational | 99 | $825M | $2B |
 NanoBanana AI - AI Image GeneratorAI-powered image generation and natural language editing |
— | Operational | |||
 CyborgAIAI avatar and video generation integration for Unity. |
Italy | Operational | |||
 KugelAudioLow-latency, multilingual text-to-speech for enterprise applications. |
Berlin, Germany | Operational | |||
 NanoBananaAI models for professional image generation and editing |
— | Operational | |||
 Z ImageEfficient open-source AI photorealistic image generation |
— | Operational | |||
 Z-Image AIOpen-source model for fast, photorealistic image generation |
— | Operational | |||
 GPT Image 1.5AI model for production-quality image generation and editing |
— | Operational |
Investments and acquisitions
Recent portfolio relationships for companies that also act as investors or acquirers.
Investors
From Dealroom's funding and investor records, grouped by the round each investor first entered.
Seed1 investor entered at this stage
Llama Venture
Source: Dealroom Talent Intelligence.
Global footprint
team presence
Market sentiment
An AI-synthesised read of the highest-engagement posts about Visko on X, ranked by likes and reposts, corporate channels excluded.
Reading the room on X — pulling top posts and synthesizing themes…
Source: X recent search ranked by engagement (likes + retweets) · Synthesis by Claude · Cached for 1 hour
See Visko’s full profile on Dealroom
Request a free trial to see the complete deal history, financials, people graph and growth signals.
Takes under a minute. Already use Dealroom? Log in.