8.0
/ 10
GPT-5.6
150 USER VOICES · REALITY SCORE
Capable frontier model with strong intent inference and generous sub limits, but HN devs report 'austere' code output and API costs remain a