GPT-5.6 Terra & Luna Price Cuts
OpenAI cut GPT-5.6 Terra to $2.00/$12.00 and Luna to $0.20/$1.20 per 1M tokens — Terra is 20% cheaper, Luna 80%. The new rates are live on LLM Gateway and apply automatically to every request, including cached input, cache writes, and long-context pricing.
Three weeks after the GPT-5.6 launch, OpenAI moved the price-performance frontier: GPT-5.6 Terra is now 20% cheaper and GPT-5.6 Luna 80% cheaper. The new rates are already live on LLM Gateway and apply to every request automatically.
New Rates
| Model | Input / 1M | Cached input / 1M | Output / 1M |
|---|---|---|---|
gpt-5.6-terra | | | |
gpt-5.6-luna | | | |
Everything derived from the base rate drops with it:
- Cache writes stay at 1.25x the input rate, so Terra writes at $2.50/M and Luna at $0.25/M.
- Long context (over 272K input tokens) stays at 2x input / 1.5x output on the new bases.
- Flex still halves the bill — Luna on flex now runs at $0.10/M input and $0.60/M output — and Priority stays at 2x.
gpt-5.6-sol pricing is unchanged at $5.00/$30.00.
Nothing to Change on Your End
The lower rates apply automatically wherever these models are billed — direct requests, Auto Route picks, and fallback traffic alike. If you route by cost, the router already sees the new prices. Capabilities are untouched: the same 1.05M-token context window, reasoning, vision, tool calling, and web search as at launch.