OpenAI's cheap, fast 5.6 variant: 147 tokens a second at $0.20 per million in.
independent benchmark score from Artificial Analysis, not Timace's own judgment
Best for
Timace is not only a directory. Every model it tracks can be dropped into something you are building.
Put GPT-5.6 Luna against up to three others on price, benchmark scores, speed and context window, with the rows they agree on hidden.
Open the comparisonPlot GPT-5.6 Luna against every other priced model and see whether the extra money buys measurably more intelligence.
The price of intelligenceWrite out the steps of a job you do repeatedly and place GPT-5.6 Luna at the step it belongs to.
Browse workflowsKeep GPT-5.6 Luna in a named list, and let the Intelligence Feed surface news about it.
Your library
