BREAKING
OpenAI Previews GPT-5.6 Series
Three Tiers: Sol, Terra, Luna
SolFlagship
Coding and science
Long-horizon agentic work
TerraBalanced
GPT-5.5-level quality
About 2x lower cost
Terminal-Bench 2.1 Scores
Sol Ultra91.9
Sol88.8
Claude Mythos 588
GPT-5.583.4
0$/M in
Input tokens
0$/M out
Output tokens
0M
Context window
Sol Refactors 15,000-Line Codebase
1Analyze 15K-line code
2Identify 47 fixes
3Implement in one session
Wider Access In Coming Weeks
AI NEWS BLITZ
OpenAI has launched a limited preview of its new GPT-5.6 model series.