OpenAI Cuts GPT-5.6 Prices — Luna Down 80%, Sol Gets a Faster Mode
Just a day after unveiling the GPT-5.6 model family on July 29, 2026, OpenAI announced further price and speed improvements. The lightweight “GPT-5.6 Luna” drops 80% in price, the balanced “GPT-5.6 Terra” drops 20%, and the flagship “GPT-5.6 Sol” gains a new “Fast mode” in the API that responds 2.5x faster. The changes apply across ChatGPT Work, Codex, and the API.
Details
- Luna price cut: Now $0.20 per million input tokens and $1.20 per million output tokens — an 80% reduction
- Terra price cut: Now $2 per million input tokens and $12 per million output tokens — a 20% reduction
- New option for Sol: Pricing itself is unchanged, but a new “Fast mode” replaces the previous “Priority Processing” in the API, delivering 2.5x the speed of standard processing at 2x the standard price
- Luna’s positioning: OpenAI says it delivers “performance comparable to models that were frontier-class a year ago” at significantly lower cost, and can still handle tool calls and multi-step workflows
- Scope: The price and speed improvements apply across ChatGPT Work, Codex, and the OpenAI API
- Goal: Make high-frequency, large-scale agentic workloads more affordable to run
How to try it
- If you’re already using GPT-5.6 Luna, Terra, or Sol via ChatGPT Work, Codex, or the OpenAI API, the new pricing applies automatically
- To get faster responses from Sol via the API, explicitly select the new “Fast mode”
- Luna remains the best value for token-heavy agentic tasks, while Terra is the balanced choice for everyday use