AI price wars: OpenAI cuts GPT-5.6 Luna prices by 80% as model competition shifts toward cost

July 30, 2026 — 2:48 pm PT

OpenAI is sharply reducing the prices of two models in its GPT-5.6 frontier series, cutting GPT-5.6 Luna (smallest/fastest) by 80% and GPT-5.6 Terra (mid-tier) by 20%, while adding a premium Fast mode for flagship GPT-5.6 Sol.

New pricing (per million tokens)

• GPT-5.6 Luna: $0.20input/$1.20 output (combined $1.40).Prior:$1 / $6(combined$7).
• GPT-5.6 Terra: $2input/$12 output (combined $14).Priorcombined$17.50.
• GPT-5.6 Sol Standard: unchanged at $5/$30.
• Sol Fast mode: $10/$60 — up to 2.5× throughput at 2× Standard price; replaces Priority Processing.

CEO Sam Altman announced the changes on X as “major price cuts.” Lower Luna/Terra prices also reflected in how usage counts against paid subscriptions for Codex and ChatGPT Work (per Business Insider / company statements cited in coverage).

Competitive context

Cuts arrive days after Anthropic released Claude Opus 5 at same price as Opus 4.8, and Google introduced Gemini 3.6 Flash and Gemini 3.5 Flash-Lite targeting lower inference costs. Luna now undercuts Gemini 3.5 Flash-Lite on combined token price and sits in the low-cost commercial tier alongside models from Xiaomi, DeepSeek, MiniMax, and others — though not the absolute cheapest on pure token basis.

VentureBeat notes Cognition and Artificial Analysis framing GPT-5.6 as strong on price/performance Pareto; Claude Opus 5 remains roughly as performant as Sol while ~6% cheaper on combined rates ($5/$25).

Positioning

OpenAI frames GPT-5.6 as frontier family: Sol for complex agentic/coding workloads; Terra for general production; Luna for high-throughput, low-latency tasks (summarization, classification, routing, lightweight assistants). Move repositions Luna into low-cost inference competition without a new model generation.