OpenAI Cuts GPT-5.6 Luna and Terra Prices

OpenAI cut API prices for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20% on July 30, leaving GPT-5.6 Sol pricing unchanged. Luna now costs $0.20 per million input tokens and $1.20 per million output tokens, while Terra costs $2 and $12 respectively. The reductions also lower how Terra and Luna usage counts against paid Codex and ChatGPT Work subscriptions.

OpenAI Cuts GPT-5.6 Luna and Terra Prices OpenAI cut API prices for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20% on July 30, while leaving GPT-5.6 Sol pricing unchanged. Luna now costs $0.20 per million input tokens and $1.20 per million output tokens, while Terra costs $2 and $12 respectively. The reductions also lower how Terra and Luna usage counts against paid Codex and ChatGPT Work subscriptions. OpenAI cut API prices for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20% on July 30, roughly three weeks after the models became generally available. GPT-5.6 Sol , the highest-capability model in the family, keeps its existing price. Luna now costs $0.20 per million input tokens and $1.20 per million output tokens , down from $1 and $6. Terra now costs $2 per million input tokens and $12 per million output tokens , down from $2.50 and $15. OpenAI's July 9 launch announcement provides the earlier rates, while its July 30 pricing post confirms the new ones. OpenAI attributed the reductions to improvements across model serving and production software. The company said GPT-5.6 Sol-assisted kernel work reduced end-to-end serving costs by 20% and that related experiments increased token-generation efficiency by more than 15%. Those are company-reported operating figures and have not been independently validated in the retrieved reporting. API and subscription changes OpenAI also introduced Fast mode for the API, replacing Priority Processing. For GPT-5.6 Sol, the company says Fast mode can deliver up to 2.5 times the speed of Standard processing at twice the price, with no change in model intelligence. Existing API requests tagged as priority continue to work and route to Fast mode. The lower Luna and Terra rates also change how usage is counted against paid subscriptions for Codex and ChatGPT Work. OpenAI said subscription prices and quota budgets remain unchanged, but use of the two models now consumes fewer credits. OpenAI published customer evaluations alongside the announcement. Notion said Terra delivered quality comparable to GPT-5.5 in its evaluations at half the cost per task and in 60% less time. Replit said Luna enabled use cases it had not expected to build soon. These are customer statements selected and published by OpenAI, not independent benchmarks. What the cuts change for teams Axios and Quartz confirmed the new rates and placed them in a market where enterprise buyers are paying closer attention to inference costs. Axios also noted competitive pressure from lower-cost Chinese open-weight models. For ML teams, the direct effect is a lower per-token bill for high-volume Luna and Terra workloads such as classification, extraction, tool use, and multi-step agents. The practical saving for any application still depends on prompt size, output length, cache reuse, retries, tool calls, latency requirements, and the quality threshold the workload must meet. The price cut therefore changes the starting economics, not the evaluation requirement. Teams comparing models should measure cost per successful task under their own traffic and quality checks rather than treating the published token rate as a complete cost benchmark. Key Points - 1OpenAI reduced Luna input pricing to $0.20 per million tokens, materially changing unit economics for high-volume API workloads. - 2Terra's 20% cut and unchanged Sol pricing create a wider cost-capability spread across the GPT-5.6 product family. - 3The cuts underscore growing emphasis on inference costs and return on investment in enterprise AI. Scoring Rationale The reductions are substantial for teams using GPT-5.6 Luna or Terra at scale, especially for token-intensive agentic and tool-use workloads. The story also provides a concrete indicator of intensifying competition around inference efficiency, though it is not a new model release or a broad platform change. Sources Primary source and supporting public references used for this report. Practice interview problems based on real data 1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with. Try 250 free problems /problems