The highest Proven Scores in AI models & APIs are NVIDIA (89/100, Proven), OpenAI (65/100, Established) and ElevenLabs (62/100, Established). They are ranked on the Proven Trust Index using Proven marketplaces, Member redemption clicks, Client quality, Operating tenure, Market standing, Connected relationships and Connected data. Inputs without data are left out and the score is scaled to 100. Vendors cannot pay to change a score.
| Rank | Name | Proven Score | Band | Best for |
|---|---|---|---|---|
| 1 | NVIDIA | 89 | Proven | Team size: 42,000 employees · Region: Santa Clara, United States · Use case: AI models & APIs |
| 2 | OpenAI | 65 | Established | Stage: Venture (Round not Specified) · Team size: 4,500 employees · Region: San Francisco, United States · Use case: AI models & APIs |
| 3 | ElevenLabs | 62 | Established | Stage: Series D · Team size: 1,500 employees · Region: London, United Kingdom · Use case: AI models & APIs |
| 4 | Anthropic | 62 | Established | Stage: Series H · Team size: 6,200 employees · Region: Cambridge, United States · Use case: AI models & APIs |
| 5 | Baseten | 47 | Emerging | Stage: Series F · Team size: 400 employees · Region: San Francisco, United States · Use case: AI models & APIs |
| 6 | Writer | 45 | Emerging | Stage: Series C · Team size: 690 employees · Region: San Francisco, United States · Use case: AI models & APIs |
| 7 | Mistral AI | 44 | Emerging | Stage: Series D · Team size: 1,200 employees · Region: Paris, France · Use case: AI models & APIs |
| 8 | Tavily | 41 | Emerging | Stage: Merger / Acquisition · Team size: 110 employees · Region: New York, United States · Use case: AI models & APIs |
| 9 | Inworld AI | 39 | Emerging | Stage: Series A · Team size: 85 employees · Region: Mountain View, United States · Use case: AI models & APIs |
| 10 | Labelbox | 38 | Emerging | Stage: Series E · Team size: 490 employees · Region: San Francisco, United States · Use case: AI models & APIs |
| 11 | fal (fal.ai) | 36 | Emerging | Stage: Series D · Team size: 180 employees · Region: San Francisco, United States · Use case: AI models & APIs |
| 12 | Groq | 33 | Emerging | Stage: Venture (Round not Specified) · Team size: 320 employees · Region: Palo Alto, United States · Use case: AI models & APIs |
| 13 | Opper AI | 32 | Emerging | Stage: Seed · Team size: 11 employees · Region: Sweden · Use case: AI models & APIs |
| 14 | Vectara | 27 | Emerging | Stage: Series A · Team size: 83 employees · Region: Palo Alto, United States · Use case: AI models & APIs |
| 15 | Cartesia | 27 | Emerging | Stage: Venture (Round not Specified) · Team size: 140 employees · Region: San Francisco, United States · Use case: AI models & APIs |
| 16 | Clarifai | 25 | Emerging | Stage: Merger / Acquisition · Team size: 47 employees · Region: Wilmington, United States · Use case: AI models & APIs |
| 17 | Twelve Labs | 25 | Emerging | Stage: Series B · Team size: 180 employees · Region: San Francisco, United States · Use case: AI models & APIs |
| 18 | Z.ai | 25 | Emerging | Team size: 990 employees · Region: Beijing, China · Use case: AI models & APIs |
| 19 | OpenRouter | 24 | Emerging | Stage: Merger / Acquisition · Team size: 120 employees · Region: New York, United States · Use case: AI models & APIs |
| 20 | xAI | 22 | Emerging. Provisional score | Team size: 12,000 employees · Region: San Jose, United States · Use case: AI models & APIs |
| 21 | Interfaze | 12 | Emerging. Provisional score | Use case: AI models & APIs |
| 22 | Palantir Technologies | 10 | Emerging. Provisional score | Use case: AI models & APIs |
| 23 | Roboflow | 10 | Emerging. Provisional score | Use case: AI models & APIs |
| 24 | Replicate | 8 | Emerging. Provisional score | Use case: AI models & APIs |
| 25 | Runware | 8 | Emerging. Provisional score | Use case: AI models & APIs |
| 26 | Anam | 6 | Emerging. Provisional score | Use case: AI models & APIs |
| 27 | Fastino Labs | 6 | Emerging. Provisional score | Use case: AI models & APIs |
Vendors in AI models & APIs are ordered by Proven Score. Data updated 2026-10-27. Inputs without data are left out and the score is scaled to 100. A score without client quality is marked Provisional score.
| Input | How it is measured | Max |
|---|---|---|
| Proven marketplaces | Investor, bank and government marketplaces on Proven that list the vendor. Log scale, full marks at 150. | 15 |
| Profile views | Member views of the vendor profile across Proven marketplaces. Log scale, full marks at 10,000. Added once the data feed is live. | 15 |
| Member redemption clicks | Clicks by verified members on the vendor's offers. Log scale, full marks at 5,000. | 15 |
| Client quality | Clients the vendor names on its own website, case studies or press releases, weighted by caliber. Full marks at 4. | 20 |
| Operating tenure | Years since founding, from a public record. Full marks at 15 years. | 10 |
| Market standing | Publicly traded or owned by a public company 10. Private company valued $1B+ or established partnership 6. | 10 |
| Connected relationships | Client relationships confirmed by CRM sync or Crossbeam. Log scale, full marks at 500. | 10 |
| Connected data | Crossbeam connected 3, CRM syncing 2. | 5 |
Vendors cannot pay to change a score. Scores can't be bought. About the data.
Source: Proven platform data, public records and vendor websites. Vendors cannot pay to change a score. Scores can't be bought. Data updated 2026-10-27. About the data · Proven