OpenAI has announced major price reductions for two GPT-5.6 API models, cutting the cost of GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20%. The changes are part of the company’s ongoing efforts to improve model efficiency and reduce the cost of using its artificial intelligence tools.
Under the updated pricing, GPT-5.6 Luna now costs $0.20 per million input tokens and $1.20 per million output tokens. The previous prices were $1 per million input tokens and $6 per million output tokens.
GPT-5.6 Terra has also received a price reduction. Input token costs have fallen from $2.50 to $2 per million tokens, while output token pricing has dropped from $15 to $12 per million tokens.
Source: OpenAI
In a post on X, OpenAI said the new pricing also affects how usage is calculated in Codex and ChatGPT Work.
When customers use the updated models for new tasks, those tasks consume fewer credits from their usage allowances. This means users may be able to complete more work within the same quota.
OpenAI is also moving Auto-review in the ChatGPT app and Codex CLI from GPT-5.4 to GPT-5.6 Luna. According to the company, the change could reduce the cost of these operations by approximately ten times.
GPT-5.6 Sol gets a faster API mode
OpenAI has introduced GPT-5.6 Sol Fast mode for API customers. The company has not announced a price reduction for Sol’s standard API access.
GPT-5.6 Sol Fast mode can process requests up to 2.5 times faster than the standard mode without reducing the model’s intelligence, OpenAI said.
The faster processing option costs twice as much as the standard GPT-5.6 Sol API price. This makes Sol Fast particularly suited to time-sensitive coding, research, and agentic AI workloads where response speed is a priority.
For general-purpose use cases that do not require faster processing, OpenAI’s standard GPT-5.6 Sol mode may be the more cost-effective option.
OpenAI said recent improvements to GPT-5.6 Sol helped deliver the efficiency gains behind the price reductions for GPT-5.6 Luna and GPT-5.6 Terra.
In the company’s internal test results, GPT-5.6 Luna ranked at the top of its intelligence index among the compared models, despite having a substantially lower cost per task.
Security teams log 54% of successful attacks and alert on just 14%. The rest move through your environment unseen.
The Picus whitepaper shows how breach and attack simulation tests your SIEM and EDR rules so threats stop slipping by detection.
Source: www.bleepingcomputer.com