How GPT-5.6 fuses frontier intelligence with frontier efficiency
The GPT‑5.6 Sol model offers enhanced efficiency and cost-effectiveness compared to previous versions, with reduced serving costs and improved performance on various benchmarks. Optimizations in routing, scheduling, and kernel management have led to a 20% reduction in costs, while the agentic harness streamlines work and reduces context bloat, further improving efficiency. These advancements reflect years of cumulative improvements across the model stack, driving future performance gains.
More from Openai
How RingCentral builds AI-native work from engineering to ops
See how RingCentral uses ChatGPT Work and Codex to accelerate AI product development and centralize operational intelligence across engineering and operations.
Daybreak models are now available on AWS
OpenAI and AWS are making Daybreak cybersecurity capabilities available through Amazon Bedrock to support enterprise security workflows.
Testing ads in ChatGPT
OpenAI begins testing ads in ChatGPT to support free access, with clear labeling, answer independence, strong privacy protections, and user control.
