GPT-5.6 Luna API Prices Double: Input Now $0.20, Output $1.20 per 1M Tokens

What Changed

OpenAI has raised GPT-5.6 Luna's API prices by +100% across both input and output. Input tokens move from $0.10 to $0.20 per 1M tokens; output tokens jump from $0.60 to $1.20 per 1M tokens. No partial increases, no tiered rollout — everything doubled.

Does It Matter for Your Workload?

That depends heavily on your token split. Most production workloads are output-heavy, so the output rate is usually the number that hurts. At $1.20 per 1M output tokens, a pipeline generating 10M output tokens per month now costs $12.00 instead of $6.00 — a straightforward $6 swing that compounds fast at scale.

For low-volume or read-heavy use cases (short completions, classification, structured extraction with brief outputs), the impact is more modest. A workload spending $20/month pre-increase lands at $40 post-increase — annoying but not a crisis.

For high-volume generative tasks — long-form content, agentic loops, coding assistants with verbose responses — the math changes meaningfully. Now is a reasonable time to benchmark alternatives.

What to Do Next

Check your last 30 days of token usage in your OpenAI dashboard. Split input vs. output, apply the new rates, and get a real number before reacting. If the delta is significant, compare against models at similar capability levels on our site.

See current rates, context limits, and how Luna's price has moved over time: GPT-5.6 Luna — live specs & price history.

A doubling is a big move. It doesn't automatically mean Luna is the wrong choice — but it does mean the cost-quality tradeoff needs a fresh look.

Ezra, Scout AI Team