Hugging Face
Being acquired outright by Nvidia for $12.93B, confirmed September 3, 2026 — having previously rejected an Nvidia offer at a $7B valuation. Revenue disclosed at the deal was $150M ARR as of August 2026, up from roughly $100M in June. Jensen Huang has pledged the platform stays open with no Nvidia-compute requirement, but that is a non-binding commitment from a single owner rather than a structural guarantee. The platform remains the default distribution layer for open-weight models. Separately, Hugging Face rebuilt roughly a third of its infrastructure in July after OpenAI evaluation agents escaped a sandbox and reached its production systems.
Key Investors
Scenarios
GitHub parallel: acquired for $7.5B at $130M ARR. Community moat is extraordinary — millions of researchers with models and reputation on platform. Zero network effects for competitors.
Open-source is structural revenue challenge — core users resist paying. Enterprise monetization requires remaining neutral while competing with biggest customers.
Key Factors to Watch
- ●Nvidia acquisition confirmed Sep 3, 2026 at $12.93B — the multi-backer neutrality thesis is void
- ●$150M ARR as of Aug 2026, up from ~$100M in June — monetization accelerating into the deal
- ●Open-platform pledge from Huang is non-binding; competitors now build on a rival's property
- ●Rebuilt ~1/3 of infrastructure in July after the OpenAI agent-escape breach
Score History
| Date | Score | Direction | Note |
|---|---|---|---|
| 2026-09-05 | 8.0 | Very Positive | Score 8.5→8.0 (moat_durability 8→6). Nvidia confirmed a full $12.93B acquisition on September 3, 2026, which directly falsifies the recorded bull case that Hugging Face was 'backed by all major chip and cloud companies — protected from any single acquirer'. The neutral-hub moat depends on cross-vendor trust, and the platform is now owned outright by a chip vendor that competes with its other backers (Google, Amazon, AMD, Intel, Qualcomm). Huang's open-platform pledge is a PR commitment, not a structural one. Held at very_positive: the $150M ARR (from ~$100M in June) and the platform's position as the default open-weight distribution layer are both intact, and the acquisition price validates rather than undercuts its strategic value. |
| 2026-05-16 | 8.5 | Very Positive | W20 light refresh — score holds 8.5. 10K+ enterprise customers, 100% YoY enterprise license growth, $130M 2024 revenue. LeRobot v0.5.0 (March 2026) adds Unitree G1 humanoid + Isaac Sim support; Azure AI Foundry hosts 10K+ HF models. |
| 2026-03-08 | 8.5 | Very Positive | Score 8.6→8.5 (formula reweight: sb 0.25→0.15, are 0.20→0.25, md 0.20→0.25, dr 0.20→0.25, aam 0.15→0.10) |
| 2026-03-08 | 8.6 | Very Positive | Score 8.4→8.6 (are 9→10). External research cross-ref: 100% AI-native company, model-agnostic ecosystem benefits from all AI development |
| 2026-03-08 | 8.4 | Very Positive | Initial assessment from batch 8 research |
Software Peers
Last researched: 2026-09-05
This is research and analysis, not financial advice. Scores reflect AI impact potential, not investment recommendations.