Uptime SLA with a figure in MLOps and model serving
"High availability" is marketing. "99.9 %" is a commitment.
Read the percentage with its scope: use the uptime calculator for a time-based example and the SLA review checklist for the questions behind it.
1 of 36 companies we measured here publish it, that is 3%. Last checked 2026-10-17.
3% here against 3% across every category we measure. MLOps and model serving sits exactly at the overall share for this criterion.
This page counts companies, not product entries: a company with two products in this category is one company here and two rows below.
Back to Uptime SLA with a figure across all categories, or to MLOps and model serving.
- Found1
- Not found at checked addresses35
- Not measured1
37 companies in total. Unmeasured companies are excluded from the published share.
Published
An availability commitment with a figure.
| Vendor | What we found |
|---|---|
| Union | https://www.union.ai/slaChecked |
No uptime figure found.
We reached at least one address for each of these vendors and none of them carried it. That is what we found at the addresses we checked, on the date we checked them, and not a statement about what the vendor has.
| Vendor | What we found |
|---|---|
| Anyscale | 6 addresses checkedChecked |
| Baseten | 6 addresses checkedChecked |
| BentoML | 6 addresses checkedChecked |
| ClearML | 6 addresses checkedChecked |
| Comet | 6 addresses checkedChecked |
| CoreWeave | 6 addresses checkedChecked |
| DataChain | 6 addresses checkedChecked |
| Dataiku | 6 addresses checkedChecked |
| DataRobot | 6 addresses checkedChecked |
| Evidently | 6 addresses checkedChecked |
| Fireworks AI | 6 addresses checkedChecked |
| Flyte | 6 addresses checkedChecked |
| Genesis Cloud | 6 addresses checkedChecked |
| H2O.ai | 6 addresses checkedChecked |
| Kedro | 6 addresses checkedChecked |
| Lambda | 6 addresses checkedChecked |
| Lightning AI | 6 addresses checkedChecked |
| Metaflow | 6 addresses checkedChecked |
| Mistral AI | 6 addresses checkedChecked |
| Modal | 6 addresses checkedChecked |
| Nebius | 6 addresses checkedChecked |
| Northern Data | 6 addresses checkedChecked |
| Nscale | 6 addresses checkedChecked |
| Paperspace | 6 addresses checkedChecked |
| Replicate | 6 addresses checkedChecked |
| RunPod | 6 addresses checkedChecked |
| Seldon | 6 addresses checkedChecked |
| StackIt | 6 addresses checkedChecked |
| Superwise | 6 addresses checkedChecked |
| Together AI | 6 addresses checkedChecked |
| Valohai | 6 addresses checkedChecked |
| Vast.ai | 6 addresses checkedChecked |
| Wallaroo | 6 addresses checkedChecked |
| Weights & Biases | 6 addresses checkedChecked |
| ZenML | 6 addresses checkedChecked |
Not measured
We could not look at these, or their robots.txt asks us not to. That says something about our measurement, not about them, and they are left out of the percentage above.
| Vendor | What we found |
|---|---|
| Domino | The vendor’s robots.txt excludes us from at least one of these addresses, so this was not measured.Checked |
Other criteria in MLOps and model serving
- Data processing agreement
- Subprocessor list
- Data location stated
- Status page with history
- Public pricing
- Certifications named
- HTTPS enforced with HSTS
- No third-party tracking before consent
- security.txt (RFC 9116)
- Notice before subprocessors change
- Named privacy contact
Uptime SLA with a figure in related categories
Found something we missed? Tell us. A correction moves that vendor into our next nightly run. That run reads the same fixed addresses we check for every vendor, not the address you send.