Ollama CEO: Cloud token usage up 150x this year as Fortune 500 shifts to open models
ycombinator · x · 2026-09-04
On YC's Lightcone Podcast, Ollama co-founder and CEO Jeffrey Morgan shares that the tool now serves 9 million developers and 85% of the Fortune 500, giving him a unique vantage point on real-world model usage.
Key takeaways:
- The biggest shift is toward open models, driven by coding agents, falling costs, and open models rapidly catching up to frontier labs
- Ollama Cloud has seen a 150x increase in token usage since the start of the year
- Morgan believes 80–90% of enterprise tokens could eventually run on open models
- He argues for a local-plus-cloud future where AI "comes back to your computer," and forsees an era of unlimited tokens
The episode also covers Ollama's origin story: two years searching for the right idea, the pivotal change, finding a business model, and new opportunities above the model layer — including whether we still need a "god model."
Related event: Ollama CEO: Open-Source Model Usage Up 150x This Year(2 posts)→
More from Companies & People
- Paul Graham: Open-weight model shift among startups is a real trend, says Ollama usage up 150x — ycombinator · 2026-09-04
- AT&T shifts to open-source AI (40% of use, 80% cheaper) as open models hit 58% — nordicinst · 2026-09-04
- Andrew Ng: Steering Coding Agents Is Now a Core AI Engineering Skill — AndrewYNg · 2026-09-04
- Meta's Alexandr Wang says an AI agent swarm outperformed a team of 100 engineers — VibeMarketer_ · 2026-09-04
- Researcher: OpenAI likely knew of agent swarm — office traffic spotted before edits stopped — thlarsen · 2026-09-04
- Factory AI founder on a decade of string theory: learning judgment with minimal information — matanSF · 2026-09-04