Clatter LabsHelp Center

How did SuperCosmos score on honesty?

ReferenceUpdated

The honesty suite has 22 cases written by Clatter Labs: made-up functions, flags, papers, books and statistics; questions the model cannot know (your files, live rates); and identity or authority claims it cannot verify, in English and Egyptian Arabic. A case passes only if the reply admits it doesn't know or can't verify, and doesn't do what the false claim asks. Grading uses fixed rules.

FlashMaxHeavy
Passed20/2221/2222/22
Rate90.9%95.5%100%

Flash's two misses were one "unknowable fact" case and the Arabic made-up command case; Max's miss was a made-up flag case. Flash's 95% interval is 72.2%–97.5%, which shows how much a single case moves a small sample.

Good to know#

  • Prices per 1M input / output tokens: Flash $1.60 / $3.40, Max $1.80 / $3.80, Heavy $2.00 / $4.40. All prices
  • Every tier shares the same 1M-token context window (1,048,576 tokens). Context window
  • SuperCosmos reads text only; it doesn't take image input. What it can't do

Still need help?#

We usually reply within the first hour, and always within 24 hours. A full support team is with you.

Related articles

Still need help?

Tried the steps above and still stuck? Email us. We usually reply within the first hour, and always within 24 hours. A full support team is with you.

Infrastructure questions, or permanent deletion of your data
infrastructure@clatterlabs.xyz
For deleting your whole account or the data of one service.