How Antinuous launched Antinuous on X

Runs large-scale experiments and simulations for AI agents on compute clusters

What Antinuous launched

AntinuousNov 19, 2026

Today, we're launching Antinuous (@ycombinator P26), a research and experimentation platform to run 100x more experiments and simulations, 1,000x cheaper. We combine swarm orchestration, compute clusters, and hardware-tuned simulation software to give every agent access to https://t.co/H7XtjwLN1c

10099706612K579

Transcript

  1. 0:00Hi, I'm Ricardo, founder of Antinous. We're building the platform to run 100 times more
  2. 0:05experiments and simulations 1,000 times cheaper. Currently, Frontier Research is accelerating at
  3. 0:10a rate we've never seen before, but the acceleration is limited to those who have
  4. 0:15access to one of three things. Unrestricted access to Frontier models, exotic lab equipment
  5. 0:21in a well-funded institution, or expensive supercomputers with clusters of high-performance
  6. 0:26compute. Frontier Research labs and AI labs have all three. Leading universities have one or two.
  7. 0:33The vast majority of people on earth have zero. We're tackling the problem in three ways.
  8. 0:38Firstly, we're building an agent purpose-built for research collaboration. Secondly, we're giving
  9. 0:44each agent access to supercomputer clusters that they can spin up and down. And finally,
  10. 0:49we're building the next generation of simulation software designed to be run by thousands of
  11. 0:54agents working in parallel. First, our agent. The Antinous agent is built from the ground up for
  12. 1:00research. It has access to every tool you would expect a Frontier Research lab to include
  13. 1:05virtually. Agent swarms can be spun up to leverage massively parallel execution to speed up its work.
  14. 1:11Every agent in a swarm can explore new ideas and chain together ideas that would take humans
  15. 1:16months to work through. Secondly, compute. Antinous agents can leverage high-performance computing
  16. 1:21resources like GPU clusters or high-speed interconnected CPU clusters like you see at a
  17. 1:27trading firm or research lab. And finally, software. Software exists for things like fluid dynamics,
  18. 1:33collisions, protein folding, genomics, backtesting, and many more. But the leading software applications
  19. 1:39for every field require seat-by-seat licenses. We think the next generation of simulations
  20. 1:45needs a stronger foundation to build on. One that is flexible so that it can be extended with code,
  21. 1:51elastic so that you can spin up thousands of instances across thousands of agents,
  22. 1:55and programmable so that you can run it to script without human involvement,
  23. 1:59speeding up the feedback loop. Our platform already runs research workflows today for
  24. 2:04researchers, analysts, and engineers at Stanford, OpenAI, NVIDIA, Morgan Stanley, and more. We're
  25. 2:10excited to open it up to the public to see what you can discover. Try it for free and give us
  26. 2:15feedback at antinous.com.

How LaunchVideo.com rated the launch

LaunchVideo.com rating

2.60
Hook
Editing / Creativity
Copy
Sentiment of launch
Distribution strategy

How the launch performed

612K
views on X
#498
of 1658 launches we track
2.9x
the 212K median launch

What this launch shows

Views for the Antinuous launch compared with median views in the directory
Views compared with directory medians
GroupViews
Antinuous, this launch612,067
Similar-sized accounts, median across 274 launches55,479
AI infrastructure / MLOps / evals launches, median across 176 launches230,126
Every launch, median across 1658 launches211,764

Antinuous reached 612K views, 2.9 times the directory median of 212K, ranking 498 of 1658 launches we track. Among AI infrastructure / MLOps / evals launches it ranks 48 of 176, 2.7 times that group's median. The post came from an account with 780 followers and reached 785 times that many people, ranking 41 of 274 launches from accounts of a similar size.

Replies ran at 1.6 per 10,000 views, well below the 3.7 median. Bookmarks came in at 9.5 per 10,000 views against a median of 14.8. One bookmark for every 1.2 likes is a high ratio and usually means people intend to come back and try the product.

The launch video runs 2:24, against a directory median of 74 seconds. It went out on a Thursday.

About the launch

Antinuous is opening a public beta of a platform that lets AI agents run experiments and simulations in parallel on compute clusters they can start and stop. It is for researchers, analysts, and engineers in applied science, engineering, academic work, and finance who need to test many variants before anything reaches a bench, a prototype, or a decision. Founder Ricardo Nunez is shipping it from San Francisco as part of Y Combinator's Spring 2026 batch. The company describes the platform as access to research primitives usually reserved for large R&D organizations. A research agent breaks a question into smaller jobs, then a swarm runs them on separate machines, reviews the results, and drops weak branches between rounds. The work can include fluid and aerodynamic studies, structural and crash analysis, circuit and battery models, CAD, lookups in databases such as UniProt and PDB, and comparisons across companies or markets. Outputs arrive as notebooks, charts, documents, or models, and each figure points back to a source or to the code that produced it. Claude, ChatGPT, Cursor, and other MCP clients can start the same runs, so an outside agent can use the cluster instead of walking through experiments one by one. In testing, the company says it produced a digital twin of an F1 car, a datacenter cooling design with an early partner, and a classifier for routing requests to a model, each within hours. The launch also says workflows are already running for researchers, analysts, and engineers at Stanford, OpenAI, NVIDIA, and Morgan Stanley. Antinuous is launching on a stated goal of 100 times more experiments and simulations at 1,000 times lower cost, and is taking public feedback on the beta.

Tags

AI & MLAI infrastructure / MLOps / evalsAI agentSeedProduct launch500K-1MB2BUSFounder-led

Ask AI about Antinuous

Questions

Frequently asked questions

The Antinuous launch post reached 612K views on X, with 706 likes, 99 reposts, 100 comments, 579 bookmarks.

Antinuous launched on X on November 19, 2026 with the message "Runs large-scale experiments and simulations for AI agents on compute clusters".

Antinuous is opening a public beta of a platform that lets AI agents run experiments and simulations in parallel on compute clusters they can start and stop.

LaunchVideo.com rated the Antinuous launch 2.60 out of 5: hook 1/5, editing / creativity 3/5, copy 4/5, sentiment of launch 4/5, distribution strategy 1/5.