Transparency alternatives to Lakera
Other LLM tooling and evaluation vendors we measure, sorted by how much we found at them and not at Lakera. This is a comparison of disclosure, not of features or fitness.
12 shown of 48 that differ from Lakera on at least one criterion we measured for both. Public transparency score, not a product or security rating.
Lakera has 8 gaps in what we could measure: Data processing agreement, Subprocessor list, Data location stated, Public pricing, Uptime SLA with a figure, No third-party tracking before consent, security.txt (RFC 9116), Notice before subprocessors change. The column “Found there, not found at Lakera” is the one that closes them.
| Vendor | Score | Found there, not found at Lakera | Found at Lakera, not found there | |
|---|---|---|---|---|
| Langfuselangfuse.comSee also Transparency alternatives to Langfuse | 74/100 | 5 Data processing agreement, Subprocessor list, Data location stated, +2 | 1 Status page with history | Compare |
| Lunarylunary.aiSee also Transparency alternatives to Lunary | 52/100 | 4 Data processing agreement, Data location stated, Public pricing, +1 | 3 Status page with history, Certifications named, Named privacy contact | Compare |
| Confident AIconfident-ai.comSee also Transparency alternatives to Confident AI | 48/100 | 3 Data processing agreement, Data location stated, Public pricing | 2 Status page with history, Certifications named | Compare |
| Openlayeropenlayer.comSee also Transparency alternatives to Openlayer | 48/100 | 3 Data processing agreement, Data location stated, Notice before subprocessors change | 2 Status page with history, Named privacy contact | Compare |
| Fiddlerfiddler.aiSee also Transparency alternatives to Fiddler | 43/100 | 3 Data processing agreement, Public pricing, Notice before subprocessors change | 2 Status page with history, Named privacy contact | Compare |
| Orqorq.aiSee also Transparency alternatives to Orq | 43/100 | 3 Data location stated, Public pricing, security.txt (RFC 9116) | 1 Status page with history | Compare |
| Braintrustbraintrust.devSee also Transparency alternatives to Braintrust | 39/100 | 3 Data processing agreement, Public pricing, Notice before subprocessors change | 3 Status page with history, Certifications named, Named privacy contact | Compare |
| Mastramastra.aiSee also Transparency alternatives to Mastra | 39/100 | 3 Data processing agreement, Public pricing, Notice before subprocessors change | 3 Status page with history, Certifications named, Named privacy contact | Compare |
| Zepgetzep.comSee also Transparency alternatives to Zep | 43/100 | 2 Subprocessor list, Public pricing | 1 Named privacy contact | Compare |
| Vectorizevectorize.ioSee also Transparency alternatives to Vectorize | 39/100 | 2 Subprocessor list, Public pricing | 2 Certifications named, Named privacy contact | Compare |
| Difydify.aiSee also Transparency alternatives to Dify | 35/100 | 2 Data location stated, Public pricing | 2 Status page with history, Certifications named | Compare |
| Galileogalileo.aiSee also Transparency alternatives to Galileo | 30/100 | 2 Data location stated, Public pricing | 3 Status page with history, Certifications named, Named privacy contact | 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.