OpenAI GPT-5.6 (July 2026): Sol, Terra, and Luna arrive for developers
OpenAI launched the GPT-5.6 model family publicly on July 9, 2026, then cut GPT-5.6 Luna pricing by 80% and Terra pricing by 20% on July 30. Here's what developers need to know.
Editorial Team
The Authority AI Tools editorial team maintains this directory using vendor documentation, dated source checks, product changelogs, and clearly identified hands-on observations where available.
OpenAI launched the GPT-5.6 model family — Sol, Terra, and Luna — announcing that all three would become publicly available on July 9, 2026 after an initial limited preview in late June. (https://openai.com/news/)
OpenAI's first unified reasoning model: 70.1% SWE-bench, 400K context, and $1.25/$10 per MTok
Looking for the right model or tool for your workflow? Browse and compare the full directory here: https://authorityaitools.com/collections/all
Three models: Luna, Terra, and Sol
GPT-5.6 comes in three tiers, from fastest to most capable:
- Luna — the fastest and most affordable member of the family, aimed at high-volume, latency-sensitive work.
- Terra — a balanced, everyday model positioned as competitive with GPT-5.5 while costing significantly less.
- Sol — the most capable model, tuned for frontier reasoning and long-horizon agentic work, with particular emphasis on domains like biology, chemistry, and cybersecurity.
OpenAI has described the family as built with strong safeguards intended to hold up under real-world adversarial pressure. (https://openai.com/news/)
How GPT-5.6 reaches developers
Beyond ChatGPT, GPT-5.6 is aimed squarely at developers. OpenAI's own Codex release notes confirm that GPT-5.6 Sol, Terra, and Luna are available on Amazon Bedrock, with first-class support for max reasoning effort, and that Bedrock model names now clearly identify the GPT-5.6 family and variant. (https://github.com/openai/codex/releases/tag/rust-v0.143.0)
That makes GPT-5.6 straightforward to adopt from the terminal via OpenAI Codex, and through the OpenAI API for application developers.
API access to GPT-5.2, Codex models, Responses API, Agents SDK, and the full OpenAI platform
Cloud coding agent with GPT-5.5 frontier model, 1M+ developers, Desktop App, in-app browser use, and parallel sandboxed environments
Reported pricing
According to launch reporting, the headline short-context rates (per 1M input / output tokens) are:
- Sol: $5 / $30
- Terra: $2.50 / $15
- Luna: $1 / $6
Terra is positioned as offering roughly GPT-5.5-class performance at a substantially lower cost, while Luna targets the cheapest, fastest tier. Confirm current pricing and context-length tiers in OpenAI's official pricing before budgeting. (https://openai.com/news/)
July 30 price reduction
OpenAI later announced that, starting July 30, 2026, GPT-5.6 Luna costs 80% less and GPT-5.6 Terra costs 20% less. The updated API rates are $0.20 per 1M input tokens and $1.20 per 1M output tokens for Luna, and $2 per 1M input tokens and $12 per 1M output tokens for Terra.
OpenAI says GPT-5.6 Terra and Luna remain available in ChatGPT Work, Codex, and the API; ChatGPT and Codex subscription prices and quota budgets remain unchanged, while usage consumes fewer credits in ChatGPT Work and Codex. The pricing update also began rolling out to AWS later on July 30. (https://openai.com/index/advancing-the-price-performance-frontier-with-gpt-5-6/)
Why it matters
A three-model split lets teams match cost to the task: reach for Sol on hard, agentic, or high-stakes reasoning; default to Terra for everyday work; and use Luna where speed and price dominate. With Bedrock and API availability confirmed via OpenAI's Codex release notes, GPT-5.6 is usable in developer workflows from day one.
Note: OpenAI's public newsroom and help center pages were not directly reachable from our automation environment at publish time. Model names, variants (Sol/Terra/Luna), Amazon Bedrock availability, and
maxreasoning-effort support are confirmed by OpenAI's official Codex release notes on GitHub; the public launch date and reported pricing come from OpenAI's July 9, 2026 announcement and launch coverage. We will update this post if official figures differ.
Sources
- OpenAI — Newsroom: https://openai.com/news/
- OpenAI — “Advancing the price-performance frontier with GPT-5.6” (July 30, 2026): https://openai.com/index/advancing-the-price-performance-frontier-with-gpt-5-6/
- OpenAI Codex release notes (GPT-5.6 Sol/Terra/Luna on Amazon Bedrock): https://github.com/openai/codex/releases/tag/rust-v0.143.0
- OpenAI Codex release notes (Bedrock names identify GPT-5.6 family/variant): https://github.com/openai/codex/releases/tag/rust-v0.144.0
Tools Mentioned in This Article
GPT-5
OpenAI's first unified reasoning model: 70.1% SWE-bench, 400K context, and $1.25/$10 per MTok
Pay-per-useOpenAI API
API access to GPT-5.2, Codex models, Responses API, Agents SDK, and the full OpenAI platform
Pay-per-useOpenAI Codex
Cloud coding agent with GPT-5.5 frontier model, 1M+ developers, Desktop App, in-app browser use, and parallel sandboxed environments
FreemiumFree Resource
2026 AI Coding Tools Comparison Chart
Side-by-side comparison of features, pricing, and capabilities for every major AI coding tool.
No spam, unsubscribe anytime.
Workflow Resources
Cookbook
Mastering OpenAI Codex CLI — Skills, MCPs & Workflows
Master OpenAI Codex CLI — agents.md skills, MCP integrations, and advanced workflows.
Cookbook
The MCP Ecosystem — Essential Servers, Setup Guides & Cross-Tool Patterns
Master the Model Context Protocol ecosystem — setup guides, essential servers, and cross-tool patterns.
Cookbook
OpenAI Codex API agent loop for implementation tasks
A repeatable API-driven loop to plan, implement, validate, and summarize coding tasks using Codex and GPT models.
MCP Server
AWS MCP Server
Interact with AWS services including S3, Lambda, CloudWatch, and ECS from your AI coding assistant.
MCP Server
Context7 MCP Server
Fetch up-to-date library documentation and code examples directly into your AI coding assistant.
MCP Server
Docker MCP Server
Manage Docker containers, images, and builds directly from your AI coding assistant.
MCP Server
Figma MCP Server
Access Figma designs, extract design tokens, and generate code from your design files.
Frequently Asked Questions
What are the GPT-5.6 models?
When was GPT-5.6 released?
Is GPT-5.6 available to developers?
Related Articles
Cursor open-sources Mixture-of-Kittens, an MoE training megakernel
Cursor Research says its open-source Mixture-of-Kittens megakernel for GB300 NVL72 systems fuses MoE communication and computation and delivered a 1.41× end-to-end tokens-per-second gain in its production stack.
Read more →Product UpdatesCursor adds Google Workspace plugins for agents
Cursor’s August 3, 2026 changelog adds Google Workspace plugins that let coding agents read, write, and act across Drive, Gmail, Calendar, Docs, Sheets, and Chat.
Read more →Product UpdatesClaude Code 2.1.224 adds self-hosted runners and cross-session messaging
Anthropic’s official Claude Code changelog update for August 7, 2026 adds self-hosted runners for web, mobile, and desktop sessions plus cross-session messaging and other workflow changes.
Read more →