OpenAI makes GPT-5.6 dramatically cheaper, putting pressure on Anthropic and rivals
Follow UsGPT-5.6 Luna gets the biggest cut, with input token pricing reduced by 80% to $0.20 per million tokens.
OpenAI also introduced a faster GPT-5.6 Sol API mode offering up to 2.5x quicker responses without changing the model's intelligence.
The revised pricing undercuts Anthropic's Claude Sonnet 4.6 and comes amid growing pressure to lower AI deployment costs for businesses.
OpenAI has announced a massive price cut for its GPT 5.6 family of AI models lowering API costs by as much as 80 pct for some offerings. This move is said to make the company’s latest AI models more affordable for developers and businesses building AI-powered applications and also intensifies the competition with rivals such as Anthropic and Chinese AI firms.
SurveyOpenAI CEO Sam Altman confirmed the changes in a post on X, saying the company aims to offer the best price/intelligence tradeoff across its AI portfolio. The decision comes as enterprises increasingly want lower AI operating costs amid rising demand for generative AI tools.
The biggest price cut applies to GPT 5.6 Luna, the lightweight model. The input token pricing has dropped by 80 per cent to $0.20 per million tokens, while output tokens now cost $1.20 per million. GPT-5.6 Terra, the company’s mid-tier model, has also become cheaper, with input token pricing reduced to $2 per million and output tokens priced at $12 per million.
Also read: Apple Q3 earnings: Record iPhone sales, weaker forecast and Tim Cook’s biggest takeaways
But unlike the Luna and Terra, GPT 5.6 Sol has not received a price cut. Instead, OpenAI has introduced a new Fast API mode which promises up to 2.5 times quicker response times while maintaining the same reasoning capabilities as the standard version. The faster variant will cost twice as much as the regular Sol model, giving developers the option to prioritise lower latency for time-sensitive workloads.
The revised pricing also narrows the gap with competitors. Anthropic’s Claude Sonnet 4.6, for instance, currently charges $3 per million input tokens and $15 per million output tokens, making OpenAI’s Terra model a more affordable option on paper.
In related news, India’s Sarvam AI has announced plans to build a one-trillion-parameter foundation model in India, launched its Epoch Builder Edition platform for India-focused AI development, announced lower AI model pricing than global rivals. The company also appointed former xAI researcher Devendra Singh Chaplot as an advisor to oversee AI research and multilingual model capabilities.
