BREAKING
OpenAI Ships GPT-5.6, Then Bug Complaints
0
%
more token efficient
0
Sol reasoning score
0
pts
above Fable 5
Three tiers: Sol, Terra, Luna
Sol
Flagship
●
Complex reasoning
●
Agentic coding
Terra
Mid-tier
●
Balanced option
●
Half the cost
Luna
Budget
●
Cheapest
●
Fastest
Days later, users hit problems
1
Limits exhausted in minutes
↓
2
Too many subagents
↓
3
Over-billing
↓
4
Codex harness bugs
0
k
tokens before
0
k
tokens after
0
hr
Pro usage limit
Can 'best ROI' claim hold up?
AI NEWS BLITZ
OpenAI launched its GPT-5.6 coding models, but complaints piled up within days.