---
title: "What is a foundation model?"
canonical: "https://supercyclehq.com/learn/foundation-model"
type: "glossary"
dataAsOf: "2026-11-02"
term: "Foundation model"
license: "Free to use and quote with attribution to supercyclehq.com. Do not redistribute as a dataset."
disclaimer: "Informational only, not investment advice. The AI Exposure Score is Supercycle's editorial opinion. Agent portfolios are simulated paper portfolios run by AI."
---

# What is a foundation model?

A foundation model is a large AI system trained on broad data that other software builds on, the way applications build on an operating system. Training one takes vast compute, data, and specialized talent, so only a small group of companies train frontier foundation models; everyone else licenses or fine-tunes them.

## Why it matters to an AI investor

The distinction between training your own model and renting someone else's separates owning the technology from using it, which is why Supercycle's Foundation Models category only includes companies that train their own. Public-market exposure to frontier labs is partial at best, since several leaders remain private.

## Example stocks (Foundation Models)

| Ticker | Name | AI Exposure |
| --- | --- | --- |
| [NVDA](https://supercyclehq.com/stock/nvidia.md) | NVIDIA | 97 (Pure Play) |
| [SOUN](https://supercyclehq.com/stock/soundhound-ai.md) | SoundHound AI | 95 (Pure Play) |
| [RXRX](https://supercyclehq.com/stock/recursion-pharmaceuticals.md) | Recursion Pharmaceuticals | 90 (Pure Play) |
| [ABSI](https://supercyclehq.com/stock/absci-corporation.md) | Absci Corporation | 90 (Pure Play) |
| [TEM](https://supercyclehq.com/stock/tempus-ai.md) | Tempus AI | 88 (Core) |

## FAQ

### Can I invest in the top AI labs directly?

Mostly not through public markets: several leading labs are private. Public exposure comes via listed companies that train their own models, hold stakes in labs, or supply them with compute and infrastructure. Each route carries different economics, and stock pages here note where the exposure actually comes from.

## Related terms

- [Training vs. inference](https://supercyclehq.com/learn/training-vs-inference.md)
- [GPU](https://supercyclehq.com/learn/gpu.md)
- [AI pure play](https://supercyclehq.com/learn/ai-pure-play.md)

---

Source: https://supercyclehq.com/learn/foundation-model. Data as of 2026-11-02. Prices are end-of-day closes, never live.

Informational only, not investment advice. The AI Exposure Score is Supercycle's editorial opinion. Agent portfolios are simulated paper portfolios run by AI.

Free to use and quote with attribution to supercyclehq.com. Do not redistribute as a dataset.