GPT-5.6: Frontier intelligence that scales with your ambition
OpenAI has launched GPT-5.6, aiming to provide more intelligence per token, stronger performance per dollar, and enhanced capabilities on demand to handle the most challenging tasks.
SummaryAI generated
OpenAI launched GPT-5.6 between July and August 2026, aiming to empower developers with better value for money. By enabling two API settings (retaining the reasoning process and achieving compression), this model increased performance by three times in the ARC-AGI-3 benchmark test, while optimizing the efficiency of the model, reasoning, and agent workflows. OpenAI also introduced two lower-cost model options, Luna and Terra, to help companies deploy large-scale AI workflows at better costs. GPT-5.6 is committed to providing more intelligence per token and stronger performance per dollar, enabling developers to plan, build, review, and test software and maximize the value of intelligence per dollar spent.
Related eventsRELATED EVENTS
- 2026-09-07 21:02The `GPT-6 Astra model was released, and OpenAI President Brockman confidently declared, “Welcome to the era of AGI.”
- 2026-09-08 19:21OpenAI released GPT-6 Astra, marking the arrival of the AGI era, with capabilities for autonomous computer operation and scientific research.
- 2026-09-08 22:40① Nvidia CEO Jensen Huang posted that OpenAI’s GPT-6 Astra, released last week, was trained using approximately 100,000 NV Link 72 clusters, and he believes that General Artificial Intelligence (AGI) has officially arrived; ② GPT-6 Astra can directly operate computers and software to perform complex tasks such as programming, reaching the most advanced level in multiple fields. OpenAI has announced the beginning of the AGI era.
Quick factsQUICK FACTS
Three timesMultiple times the score improvement
TwoNumber of API settings
Key entitiesKEY ENTITIES
GPT-5.6KiroLunaOpenAITerra
Event frameEVENT FRAME
Launch
企业 · 官方发布across 5 days
Status
GPT‑5.6 and Luna/Terra solutions have been released. Developers are optimizing their workflows using dual settings and low-cost models.
Coverage · reports per dayLANGUAGE SPLIT
Coverage
Entity relations
Integrated timelineUNIFIED TIMELINE
-
2026-07-09
GPT‑5.6 is officially launched
OpenAI launches GPT‑5.6, aiming to provide more intelligence per token, stronger performance per dollar, and enhanced capabilities on demand.
-
2026-07-29
Dual settings improve benchmark scores
GPT‑5.6 passes the ARC‑AGI‑3 benchmark by enabling two API settings, triple the score improvement, while optimizing model and inference efficiency.
2 reports -
2026-07-30
Low-cost Luna/Terra solutions are launched
OpenAI offers lower-priced versions of Luna and Terra models, helping businesses deploy large-scale AI workflows at better costs.
-
2026-08-13
GPT‑5.6 developer guide is available
Startups use GPT‑5.6 to build faster, more cost-effective AI agents, optimizing processes through intelligent model selection and the Responses API.
-
2026-08-24
GPT‑5.6 is released on the Kiro platform
Kiro launches GPT‑5.6, helping developers plan, build, review, and test software at better price-performance ratios.
SignalsSIGNALS
Keyword heat
All reports (6)SOURCES
How GPT-5.6 fuses frontier intelligence with frontier efficiency
GPT-5.6 improves the AI efficiency in model, inference, and agent workflows, enabling the delivery of more useful intelligence at lower costs. This version achieves this goal through cross-model optimization, aiming to increase the value of intelligence produced per dollar spent.
How enabling two settings tripled our scores on the ARC-AGI-3 benchmark
GPT-5.6’s score in the ARC-AGI-3 benchmark was tripled by enabling two API settings. This improvement was achieved by preserving the reasoning process and implementing model compression, while also enhancing performance and efficiency.
Advancing the price-performance frontier with GPT-5.6
OpenAI has introduced lower-price options for the Luna and Terra models, aiming to help businesses deploy large-scale AI workflows more efficiently. This initiative relies on more efficient model technologies such as GPT-5.6, striving to push the boundaries between price and performance forward, enabling businesses to obtain advanced computing power at a better cost.
The builder’s guide to GPT‑5.6
Startups use GPT-5.6 to build faster and more cost-effective AI agents, through smarter model selection and the new Responses API functionality. This guide covers how to utilize these technical features to optimize the development process.
Advancing price-performance for developers with GPT‑5.6 in Kiro
Kiro has now launched GPT‑5.6, which helps developers plan, build, review, and test software with better price-performance ratios. This model is designed to improve development efficiency and optimize cost-effectiveness.