GPT-5.6 Luna Pro API Prices Cut 50%: Input Now $0.50, Output $3.00 per 1M Tokens

What Changed

OpenAI has cut GPT-5.6 Luna Pro — live specs & price history pricing by 50% across the board. Input tokens drop from $1.00 to $0.50 per 1M tokens; output tokens fall from $6.00 to $3.00 per 1M tokens.

Does It Matter for Your Workload?

The output cut is the one to pay attention to. Most real-world workloads — summarisation, code generation, agentic chains — are output-heavy, so a -50% move on output pricing has an outsized effect on total API spend.

A quick sanity check: if you were previously burning $600/month on output tokens alone (100M output tokens at $6.00), that same volume now costs $300. Savings scale linearly, so high-volume users feel this immediately.

For lighter, input-dominated use cases — classification, retrieval re-ranking, embeddings-adjacent tasks — the input drop from $1.00 to $0.50 still helps, but the absolute dollar impact is smaller.

Competitive Context

Price cuts at this magnitude typically signal either improved inference efficiency on OpenAI's end or competitive pressure from alternatives. Either way, the result is the same: Luna Pro becomes more defensible against mid-tier models that previously undercut it on cost. If you were routing certain tasks away from Luna Pro purely on price, that calculus is worth revisiting.

Bottom Line

No change to capabilities, rate limits, or context window — this is a straight price reduction. Audit your current token split (input vs. output), run the numbers at the new rates, and compare against alternatives on this site before assuming Luna Pro is now the automatic winner for your stack.

Ezra, Scout AI Team