changelogOpenAI

OpenAI Cuts GPT-5.6 Terra Price 20%, Luna Price 80% Across API and ChatGPT

TL;DR

OpenAI is cutting API prices for its GPT-5.6 Terra and Luna models by 20% and 80%, respectively, compared to prices set earlier this month. The company says the lower costs are also reflected in usage limits for ChatGPT Work and Codex subscribers, though subscription prices remain unchanged.

2 min read
0

OpenAI is cutting prices on two of its three GPT-5.6 models today, reducing API costs for Terra by 20% and for Luna by 80% compared to the initial pricing set earlier this month, according to OpenAI.

The GPT-5.6 lineup consists of three models:

  • Sol — OpenAI's flagship model
  • Terra — the balanced model for everyday work
  • Luna — the fast, affordable model

New API Pricing

According to OpenAI, the updated per-million-token rates are:

  • Terra: $2 input / $12 output (down from previous pricing, a 20% cut)
  • Luna: $0.20 input / $1.20 output (down from previous pricing, an 80% cut)

Sol's pricing was not mentioned in OpenAI's announcement and appears unchanged.

What Changes for ChatGPT and Codex Users

The price cuts aren't limited to developers calling the API directly. OpenAI says the lower costs for Terra and Luna are now reflected in how usage is counted against paid ChatGPT subscriptions.

"These lower prices for Luna and Terra are also reflected in how usage is counted against paid subscriptions when using Codex and ChatGPT Work," OpenAI said in its announcement.

Monthly subscription prices for ChatGPT Work and Codex remain unchanged, according to OpenAI. Instead, the adjustment means subscribers get more usage of Terra and Luna before hitting their plan limits, since each request now consumes a smaller share of allotted usage credits.

OpenAI did not disclose exact figures for how much additional usage this translates to for subscribers, nor did it specify whether the change applies to other ChatGPT tiers beyond Work.

What this means

This is a usage-economics change, not a new model release. By cutting Luna's cost 80% and Terra's by 20%, OpenAI is making its mid-tier and budget GPT-5.6 models cheap enough to serve as default choices for high-volume tasks — coding loops in Codex, routine ChatGPT Work queries, and API-driven applications that previously had to ration calls to more expensive models.

The fact that the discount flows through to subscription usage limits, rather than just API pricing, suggests OpenAI is trying to reduce complaints about hitting caps on paid plans without cutting subscription prices outright. It's a cheaper way to increase perceived value: instead of lowering the $ figure on a monthly bill, OpenAI stretches how far that bill goes.

For developers, the Luna price cut is the more significant move — an 80% reduction pushes it toward commodity pricing territory, positioning it against low-cost models from competitors for simple, high-throughput tasks. Terra's smaller 20% cut suggests OpenAI sees less competitive pressure on its mid-tier offering. Expect rivals to respond with their own pricing adjustments on comparable fast/cheap model tiers in the coming weeks.

Related Articles

product update

OpenAI Launches Agents API in Public Beta, Exposing Codex Infrastructure to Developers

OpenAI has released the Agents API in public beta, giving developers access to the same cloud infrastructure that powers Codex and ChatGPT. The API supports long-running agents, parallel tool use, and sub-agent delegation, with billing based solely on token usage.

benchmark

AWS Benchmark: OpenAI's GPT-5.6 Luna Beats GPT-5.4 Mini on Cost-Per-Correct-Answer Despite Similar List Price

An AWS blog post using an open-source benchmarking harness finds that GPT-5.6 Luna, Terra, and Sol on Amazon Bedrock deliver lower cost-per-correct-answer than OpenAI's cost-optimized GPT-5.4 Mini and Nano, once accuracy, token efficiency, and agent turn counts are factored in. The analysis also cites a July 30, 2026 price cut of up to 80% for GPT-5.6 Luna on Amazon Bedrock.

model release

OpenAI Launches GPT-Live-1 API for Full-Duplex Voice Apps That Talk and Listen Simultaneously

OpenAI has released GPT-Live-1 as a developer API, a speech model capable of full-duplex conversation—listening and talking simultaneously. It already powers ChatGPT's voice mode and costs $0.05 per minute, with benchmark scores showing sharp improvements over GPT-Realtime-2.1.

product update

OpenAI Launches ChatGPT for Financial Services to Automate Wall Street Analyst Work

OpenAI launched ChatGPT for Financial Services, a tailored enterprise product built with design partners Morgan Stanley and Evercore that automates research, financial analysis, and pitchbook creation. The tool, powered by GPT-6 Astra, targets tasks traditionally performed by Wall Street's junior analysts and associates.

Comments

Loading...