Announcing a major Price drop for 5.6 Terra and Luna and Fast mode for 5.6-Sol

Announcements
, ,
1

OpenAI tasked GPT-5.6 with optimizing its own runtime efficiency.

The results:

  • 20% lower serving costs through production GPU kernel improvements.
  • More than 15% better token-generation efficiency through improved speculative decoding.

These improvements are being passed on to everyone using the API, Codex, and ChatGPT.

Starting today, GPT-5.6 Luna, the fastest and most affordable model, will cost 80% less. GPT-5.6 Terra, the balanced model for everyday work, will cost 20% less.

HOfepUra4AAUEZi
HOfepUra4AAUEZi680×531 24.3 KB

Along the price reductions for GPT-5.6 Luna and Terra, Fast mode for GPT-5.6 Sol in the API delivers up to 2.5x the speed of Standard processing at twice the Standard price. It gives API customers faster access to GPT-5.6 Sol without changing the model’s intelligence.

OpenAI is also upgrading Auto-review in the ChatGPT app and Codex CLI from GPT-5.4 to GPT-5.6 Luna. Combined with Luna’s new pricing, Auto-review is expected to cost about 10x less. This affects auto-approve mode (“review for me” in the app), which now uses Luna to prevent many high risk actions from the main agent

3

nice, with the new prices gpt-5.6-luna beats gpt-5.4-mini in pricing and consolidates as the best option for cheaper runs.

4

Note: fast on API == "service_tier": "priority"

GPT-5.6 API pricing changes

Prices are in USD per 1 million tokens.

Model Input — Old Input — New Cached — Old Cached — New Output — Old Output — New
GPT-5.6 Sol $5.00 $5.00 $0.50 $0.50 $30.00 $30.00
GPT-5.6 Terra $2.50 $2.00 $0.25 $0.20 $15.00 $12.00
GPT-5.6 Luna $1.00 $0.20 $0.10 $0.02 $6.00 $1.20

Image cost report

by API calls just done

Vision: not hit with a humongous token cost multiplier on Luna (yet?) for input images as with other mini or nano models - just an unpublished-anywhere multiplication factor: 120% the cost for image tokens, across all gpt-5.6 models, that you’d arrive at after otherwise following the vision documentation exactly.

note: Up to 1500 individual image inputs per request

Updated (with model cost truth table in code): OpenAI Vision Token Calculator