GPT-6 Sol and Luna push the cost efficiency frontier by halving cost relative to GPT-5.6 Sol and Luna. Intelligence Index and Coding Agent Index scores remain level with GPT-5.6, with progress in some evaluations and regressions in others
Pricing is approximately half that of GPT-5.6: Sol drops from $4/$20 to $2/$10 per million input/output tokens, and Luna from $0.20/$1.20 to $0.10/$0.50, with the same 90% discount for cache reads and 25% premium for cache writes.
Key takeaways:
➤ Halves Cost per Task: GPT-6 Sol (max) costs $1.06 per task to run the Artificial Analysis Intelligence Index, ~50% less than GPT-5.6 Sol (max) at $1.99. GPT-6 Luna (max) costs $0.07 per task, ~60% less than GPT-5.6 Luna (max) at $0.18. This is driven by the price cut, as both models use slightly more output tokens per task (31k vs 29k for Sol, and 51k vs 41k for Luna). These two releases allow OpenAI to capture a significant portion of the cost efficiency Pareto frontier.
➤ In the Coding Agent Index, Sol improves but Luna regresses: In OpenAI's Codex harness, GPT-6 Sol (max) scores 57 in the Artificial Analysis Coding Agent Index, up 2 points from GPT-5.6 Sol (max), with gains in Terminal-Bench 4.0…
GPT-6 Sol (max) costs $1.06 per task on Artificial Analysis's Intelligence Index, about half GPT-5.6 Sol's $1.99, driven mainly by the price cut since both models use slightly more output tokenThe chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.Full definition → per task.
On the Coding AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → Index, GPT-6 Sol improves 2 points to 57 (gains on Terminal-Bench 4.0 and SWE-Atlas-QnA) while GPT-6 Luna regresses 2 points to 41, with lower scores on SWE-Atlas-QnA and DeepSWE v1.1.
Both models hallucinationWhen a model states something false with full confidence — inventing facts, citations, or APIs that don't exist.Full definition → less on Artificial Analysis's Omniscience benchmarkA standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.Full definition →: Sol's rate falls from 92% to 60%, partly by answering fewer questions (83% vs 99% attempt rate), trading some accuracy for fewer wrong answers.
GPT-6 Sol drops about 100 Elo points on GDPval-AA v2.1 (tasks across 44 occupations) and Luna drops roughly 75; Artificial Analysis's manual review attributes this to weaker presentation quality and missed rubric elements.
Pricing halves versus GPT-5.6: Sol falls from $4/$20 to $2/$10 per million input/output tokens and Luna from $0.20/$1.20 to $0.10/$0.50, keeping the same 90% cache-read discount and 25% cache-write premium.
Terms in this piece · Glossary
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
hallucination — When a model states something false with full confidence — inventing facts, citations, or APIs that don't exist.
benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.
Read the full post on X ↗
