Source: Veridrop · Tested host: pqh.ai
- Median score
- 95 · Excellent
- Tests in total
- 54
- Last test
- Days with tests in the last 90 days
- 20
- Protocol
- Claude, Gemini, OpenAI
Pass rate by check
| Check | Protocol | Passed |
|---|---|---|
| Behaviour pattern check | Claude | 33 of 39 |
| Model field and output stability check | Claude | 30 of 40 |
| Self-reported identity check | Claude | 33 of 40 |
| Streaming consistency check | Claude | 38 of 40 |
| Message ID format check | Claude | 40 of 40 |
| Protocol field check | Claude | 40 of 40 |
| Thinking signature check | Claude | 32 of 40 |
| Usage figure consistency check | Claude | 30 of 40 |
| Basic request check | Gemini | 2 of 2 |
| Function call check | Gemini | 2 of 2 |
| Streaming consistency check | Gemini | 2 of 2 |
| Model field and output stability check | Gemini | 2 of 2 |
| Protocol field check | Gemini | 2 of 2 |
| Structured output check | Gemini | 2 of 2 |
| Usage figure consistency check | Gemini | 2 of 2 |
| Basic request check | OpenAI | 12 of 12 |
| Function call check | OpenAI | 11 of 12 |
| Streaming consistency check | OpenAI | 9 of 12 |
| Model field and output stability check | OpenAI | 10 of 12 |
| Protocol field check | OpenAI | 11 of 12 |
| Structured output check | OpenAI | 11 of 12 |
| Usage figure consistency check | OpenAI | 8 of 12 |
- Counting window
- —
- Score standard deviation
- 16.2
- Reports with serious issues
- 1 of 54