Source: Veridrop · Tested host: ideastead.com
- Median score
- 97 · Excellent
- Tests in total
- 37
- Last test
- Days with tests in the last 90 days
- 11
- Protocol
- Claude, Gemini, OpenAI
Pass rate by check
| Check | Protocol | Passed |
|---|---|---|
| Behaviour pattern check | Claude | 10 of 10 |
| Model field and output stability check | Claude | 13 of 13 |
| Self-reported identity check | Claude | 13 of 13 |
| Streaming consistency check | Claude | 13 of 13 |
| Message ID format check | Claude | 13 of 13 |
| Protocol field check | Claude | 13 of 13 |
| Thinking signature check | Claude | 9 of 9 |
| Usage figure consistency check | Claude | 9 of 13 |
| Basic request check | Gemini | 5 of 5 |
| Function call check | Gemini | 5 of 5 |
| Streaming consistency check | Gemini | 5 of 5 |
| Model field and output stability check | Gemini | 5 of 5 |
| Protocol field check | Gemini | 5 of 5 |
| Structured output check | Gemini | 5 of 5 |
| Usage figure consistency check | Gemini | 4 of 5 |
| Basic request check | OpenAI | 19 of 19 |
| Function call check | OpenAI | 19 of 19 |
| Streaming consistency check | OpenAI | 19 of 19 |
| Model field and output stability check | OpenAI | 19 of 19 |
| Protocol field check | OpenAI | 19 of 19 |
| Structured output check | OpenAI | 19 of 19 |
| Usage figure consistency check | OpenAI | 6 of 19 |
- Counting window
- —
- Score standard deviation
- 4.2
- Reports with serious issues
- 0 of 37