Source: Veridrop · Tested host: 91token.com
- Median score
- 96 · Excellent
- Tests in total
- 129
- Last test
- Days with tests in the last 90 days
- 22
- Protocol
- Claude, Gemini, OpenAI
Pass rate by check
| Check | Protocol | Passed |
|---|---|---|
| Behaviour pattern check | Claude | 68 of 69 |
| Model field and output stability check | Claude | 70 of 76 |
| Self-reported identity check | Claude | 72 of 76 |
| Streaming consistency check | Claude | 67 of 75 |
| Message ID format check | Claude | 76 of 76 |
| Protocol field check | Claude | 76 of 76 |
| Thinking signature check | Claude | 63 of 76 |
| Usage figure consistency check | Claude | 48 of 75 |
| Basic request check | Gemini | 9 of 9 |
| Function call check | Gemini | 8 of 9 |
| Streaming consistency check | Gemini | 8 of 9 |
| Model field and output stability check | Gemini | 8 of 9 |
| Protocol field check | Gemini | 9 of 9 |
| Structured output check | Gemini | 9 of 9 |
| Usage figure consistency check | Gemini | 9 of 9 |
| Basic request check | OpenAI | 44 of 44 |
| Function call check | OpenAI | 43 of 43 |
| Streaming consistency check | OpenAI | 42 of 43 |
| Model field and output stability check | OpenAI | 40 of 44 |
| Protocol field check | OpenAI | 42 of 44 |
| Structured output check | OpenAI | 43 of 43 |
| Usage figure consistency check | OpenAI | 14 of 43 |
- Counting window
- —
- Score standard deviation
- 8.2
- Reports with serious issues
- 1 of 129