Cognition Updates FrontierCode 1.1 With GPT-5.6 Discounts
Benchmark revision adds discounts for GPT-5.6 models to highlight efficiency gains.
Cognition revised its FrontierCode 1.1 benchmark to incorporate new price discounts for GPT-5.6 Terra and Luna models. The update positions the GPT-5.6 series on the Pareto curve of price/performance efficiency. Commenters observed the resulting price/performance shift and noted impacts on other models in the benchmark. All Devin users will see the cost decreases reflected automatically. The leaderboard at cognition.com/frontiercode shows updated scores, methodology, and sample tasks for frontier AI models on production-quality code generation.
We've updated FrontierCode 1.1 to reflect new discounts for GPT-5.6 Terra and GPT-5.6 Luna. With these new costs, the GPT-5.6 series sits on the pareto curve of price/performance efficiency.
Reactions from ranked influencers
9 postsCheck out all model scores, methodology, and sample tasks:
GPT-5.6 series has best price/performance
We've updated FrontierCode 1.1 to reflect new discounts for GPT-5.6 Terra and GPT-5.6 Luna. With these new costs, the GPT-5.6 series sits on the pareto curve of price/performance efficiency.
this should give you an idea of what's coming in the next few months frontiermogging
We've updated FrontierCode 1.1 to reflect new discounts for GPT-5.6 Terra and GPT-5.6 Luna. With these new costs, the GPT-5.6 series sits on the pareto curve of price/performance efficiency.
Incredible price/performance shift with new pricing of Luna and Terra
We've updated FrontierCode 1.1 to reflect new discounts for GPT-5.6 Terra and GPT-5.6 Luna. With these new costs, the GPT-5.6 series sits on the pareto curve of price/performance efficiency.
@scaling01 I love capitalism
Chinese models, dead Sonnet, dead Composer, dead Gemini Flash, dead
We've updated FrontierCode 1.1 to reflect new discounts for GPT-5.6 Terra and GPT-5.6 Luna. With these new costs, the GPT-5.6 series sits on the pareto curve of price/performance efficiency.