← Scoreboard
Kimi K3
Each provider gets the same short prompts. We count how often its top answer is the same as Moonshot AI's own.
Tested
10
of 29 on OpenRouter
Match
8
Watching
2
0 no match
Typical agreement
65%
same answer as the maker
| Provider | Same answer as the maker · typical | Result | Precision | Price out / 1M |
|---|---|---|---|---|
| Moonshot AI | makersame answer twice 71% of the time | Reference | unknown | $15.00 |
| InferenceNet · fp4 | 79%84 of 106 prompts · typical 63% | Matches | fp4 | $15.00 |
| InferenceNet · us | 79%84 of 106 prompts · typical 63% | Matches | fp4 | $15.00 |
| Decart · mxfp4 | 65%69 of 106 prompts · typical 63% | Matches | mxfp4 | $9.75 |
| Chutes · mxfp4 | 65%69 of 106 prompts · typical 63% | Matches | mxfp4 | $9.50 |
| Sail Research · fp4 | 63%67 of 106 prompts · typical 65% | Matches | fp4 | $13.26 |
| Relace · fp4 | 63%67 of 106 prompts · typical 65% | Matches | fp4 | $9.34 |
| Sail Research · fp4 us | 63%67 of 106 prompts · typical 65% | Matches | fp4 | $13.26 |
| Phalaoffline | 39%7 of 18 prompts · typical 72% | On watch | unknown | $12.75 |
| Wafer · us | 37%39 of 106 prompts · typical 65% | On watch | mxfp4 | $9.60 |
Question by question
106 questions. Each square is one question. Lit means the same answer as Moonshot AI. Hover or tap a column.
Results from Nov 11. Short prompts at temperature 0, so settings cannot change the answer. Honest providers still differ a little, so each one is compared with what is typical for the others. Quality checks are not running yet. A provider is marked as no match only after 2 failed checks on different days.