Alternatives to Modal
Other MLOps and model serving vendors we measure, sorted by how much they publish that Modal does not. This is a comparison of disclosure, not of features or fitness.
12 shown of 25 that differ from Modal on at least one criterion we measured for both. Public transparency score, not a product or security rating.
Modal has 8 gaps in what we could measure: Subprocessor list, Data location stated, Status page with history, Certifications named, Uptime SLA with a figure, No third-party tracking before consent, security.txt (RFC 9116), Named privacy contact. The column “publishes and Modal does not” is the one that closes them.
| Vendor | Score | Publishes, and Modal does not | Modal publishes, and they do not | |
|---|---|---|---|---|
| Nebiusnebius.com | 30/100 | 4 Status page with history, Certifications named, security.txt (RFC 9116), +1 | 3 Data processing agreement, Public pricing, Notice before subprocessors change | Compare |
| Anyscaleanyscale.com | 61/100 | 3 Data location stated, Certifications named, Named privacy contact | nothing | Compare |
| DataRobotdatarobot.com | 43/100 | 3 Data location stated, Status page with history, Named privacy contact | 2 Data processing agreement, Notice before subprocessors change | Compare |
| Paperspacepaperspace.com | 43/100 | 3 Data location stated, Status page with history, Certifications named | 2 Data processing agreement, Notice before subprocessors change | Compare |
| CoreWeavecoreweave.com | 35/100 | 3 Data location stated, Status page with history, Certifications named | 3 Data processing agreement, HTTPS enforced with HSTS, Notice before subprocessors change | Compare |
| Northern Datanortherndata.de | 30/100 | 3 Data location stated, security.txt (RFC 9116), Named privacy contact | 3 Data processing agreement, Public pricing, Notice before subprocessors change | Compare |
| Basetenbaseten.co | 52/100 | 2 Status page with history, Named privacy contact | nothing | Compare |
| H2O.aih2o.ai | 39/100 | 2 Certifications named, Named privacy contact | 1 Public pricing | Compare |
| Nscalenscale.com | 39/100 | 2 Subprocessor list, Status page with history | 2 Data processing agreement, Public pricing | Compare |
| BentoMLbentoml.com | 32/100 | 2 Data location stated, Status page with history | 3 Data processing agreement, Public pricing, Notice before subprocessors change | Compare |
| Cometcomet.com | 30/100 | 2 Status page with history, Certifications named | 2 Data processing agreement, Notice before subprocessors change | Compare |
| Lambdalambdalabs.com | 17/100 | 2 security.txt (RFC 9116), Named privacy contact | 3 Data processing agreement, Public pricing, Notice before subprocessors change | Compare |
Only criteria we could measure for both are counted. Where we could measure one and not the other, the row is silent: that would be a statement about our measurement, not about either vendor. Every single result with its source is on the two profiles.