OptionProbability
Costs at least $50 per million output tokens on release
Scores at least 40% on Terminal Bench Science
Scores at least 55% on Terminal Bench 4
It performs worse than GPT-5.6 Sol on at least one benchmark OpenAI reports in their release blog post or system card
OpenAI claims it is AGI within one week of release
Resolves a FrontierMath Open Problems "Solid Result" or Higher Within One Month of Release
Scores at least 67 on Artificial Analysis Intelligence Index
Scores at least 50% on Gene Bench Pro
Scores at least 80% on DeepSWE
Scores at least 60% on Exploit Gym
100
98
94
0
OptionVotes
NO
YES
2511
398
1062
942
Haiku 5.5 gets 35+ on AA
Fable 5.5 is released in 2026
sonnet 5.5 ranks worse than sonnet 5 on at least 1 benchmark in the system card
sonnet 5.5 become only accessable on paid plans
Sonnet 5.5 ranks above gpt 5.6 sol on terminal bench 4.0
claude hands out a weekly reset when they drop their new model
Sonnet 5.5 is released in September
sonnet or haiku 5.5 are dropped on a weekend
67
50
40
24
12
10
