Series A
We pull jobs from sources most job seekers never check
About the company
What they build and who they serve.
Standard Intelligence is a San Francisco-based AI research lab building foundation models optimized for computer use — agents that interact with software through its graphical interface. Its FDM-1 model is trained on video footage rather than annotated screenshots, using an inverse dynamics model to auto-generate captions, and ships with a video encoder the team says is roughly 100x more efficient than comparable alternatives. The six-person team is pursuing what it calls aligned artificial general intelligence by scaling agentic capabilities through video pretraining.
Why it matters
Signals from funding, traction, and product direction.
- Round priced at a $500M post-money valuation per Techmeme/The Information, despite the company being a six-person team in San Francisco.
- Sequoia partners Sonya Huang and Mikowai Ashwill, plus Spark Capital partner Yasmin Razavi, joined the round as deal partners.
- Angel and advisor cohort includes Andrej Karpathy, Stanley Druckenmiller, and Tesla Optimus engineering lead Milan Kovac.
- Standard Intelligence assembled an 11-million-hour computer-use training dataset for FDM-1 — multiple orders of magnitude larger than the next-best open-source alternative.
Funding history
Rounds and investors over time, newest first.
- $75MSequoia Capital · LeadSpark Capital · Lead
In the news
Press, announcements, and coverage about this company.
Investors
Focus areas
No open roles on the board yet
We do not have indexed listings for Standard Intelligence right now. Roles may still be open on the company site—we will add them here when they appear in our feed.

Don’t get beat to tomorrow’s openings
New roles go live every minute and the earliest applicants win. Get the freshest, verified listings delivered straight to your inbox before most job seekers ever see them.
👉 Get free daily job posts