GPT-6 Sol costs $2/1M input and $10/1M output tokens, and GPT-6 Luna costs $0.10/1M input and $0.50/1M output tokens, both about 50% cheaper than GPT-5.6

Your AI assistant costs are cut by up to half, making more powerful AI accessible for everyday use.

OpenAI has released GPT-6 Sol and GPT-6 Luna, new versions of its advanced language models, priced approximately 50% lower than their predecessors. GPT-6 Sol costs $2 per million input tokens and $10 per million output tokens, a reduction from GPT-5.6 Sol's $4 and $20 rates. GPT-6 Luna is now priced at $0.10 per million input tokens and $0.50 per million output tokens, down from $0.20 and $1.20. These models were trained using methods similar to the high-performing Astra model, incorporating improvements in factuality, coding, computer use, and alignment. Despite Astra remaining OpenAI's top model, GPT-6 Sol demonstrates superior performance on benchmarks like AutomationBench and Agents' Last Exam compared to previous OpenAI models and Anthropic's Claude Opus 5. GPT-6 Sol and Luna also feature enhanced communication clarity and alignment improvements, exhibiting lower rates of policy circumvention. The new models are available to various tiers of ChatGPT and Codex users, with GPT-6 Luna accessible to free users on the desktop app.

OpenAI's strategic pricing of GPT-6 Sol and Luna indicates a clear response to market demand for more cost-effective AI solutions, particularly for enterprise-level applications and agentic deployments that consume substantial token volumes. By retrofitting their state-of-the-art models for efficiency and reducing prices, OpenAI aims to maintain its market leadership against emerging competitors like Jev, AlohaJet, and Pathway, which are exploring alternative scaling approaches. This move suggests a broader industry trend toward balancing model capability with affordability, potentially accelerating the adoption of advanced AI across a wider range of business functions.

The introduction of these more affordable, yet highly capable, models signals a potential shift in how AI services are consumed, moving beyond premium-only access for top-tier performance. This pricing strategy may pressure competitors to either match OpenAI's cost reductions or differentiate through specialized capabilities. The improved alignment and communication features also suggest a focus on user experience and safety, which could become key competitive factors as AI systems become more integrated into daily workflows.