$ cat models/deepseek-v4.1-flash
DeepSeek V4.1 Flash
DeepSeek
Check that an endpoint selling DeepSeek V4.1 Flash really serves it. A test compares the endpoint's answers with answers collected from DeepSeek's own API.
- Official Provider
- DeepSeek's own API
- Reference Answers
- 640
- Declared Precision
- not disclosed
- Official Release
- fp8, 32x32 blocks, experts fp4
- Checks
- Identity · Competence · Refusals (3 probes)
- Reference Data Updated
- 2026-10-16
How Well a Test Works
Size
Requests
Genuine Confirmed
Wrongly Flagged
Substitutes Caught
Standard
80
99.5%
0.5%
100.0%
Thorough
160
99.5%
0.5%
100.0%
Simulated from the reference answers. A substitute that isn't caught ends inconclusive: it is never confirmed as this model.