Overview
ChatGPT is OpenAI's AI assistant for writing, analysis, coding, research, documents, presentations, spreadsheets, and tool-using agent workflows. Its current model family in this listing is GPT-5.6, released on July 9, 2026 across ChatGPT, Codex, and the OpenAI API.
The important change is that ChatGPT is no longer a single model choice. GPT-5.6 is split into three capability tiers: Sol for the hardest work, Terra for balanced everyday capability, and Luna for faster, lower-cost execution. That lets teams choose whether a task needs maximum reasoning, cheaper high-volume processing, or a balance between the two.
For individuals, ChatGPT remains a general-purpose assistant. For teams, it is increasingly a work system: ChatGPT handles conversations and artifacts, Codex handles software tasks, and the API lets developers build workflows with Programmatic Tool Calling, explicit cache breakpoints, and beta multi-agent orchestration.
Key Features
GPT-5.6 Sol, Terra, and Luna - Sol is OpenAI's flagship GPT-5.6 tier, Terra is the balanced tier, and Luna is the fastest and lowest listed API-price tier. The structure helps teams avoid using the most expensive reasoning path for every task.
ChatGPT Work and Codex access - GPT-5.6 is available across ChatGPT Work and Codex with plan-dependent model selection and effort controls. This makes ChatGPT more useful for app building, code review, debugging, design iteration, and long-running technical work.
API support for agent workflows - Developers can access GPT-5.6 through the OpenAI API. The Responses API adds Programmatic Tool Calling and beta multi-agent support, giving teams a more structured path for tool coordination than a plain chat loop.
Knowledge-work output quality - GPT-5.6 is positioned for documents, presentations, spreadsheets, research, and design-heavy outputs. That matters for teams using ChatGPT to produce shareable work rather than only answers.
Prompt caching controls - GPT-5.6 introduces more predictable cache behavior, including explicit cache breakpoints and a 30-minute minimum cache life. This is useful for repeated long prompts, agent scaffolds, and retrieval-heavy workflows.
Broad everyday assistant coverage - ChatGPT still covers drafting, summarization, brainstorming, translation, data analysis, image understanding, and coding help for general users, with advanced access depending on plan and rollout.
Pricing & Plans
ChatGPT has free and paid plans, with model access varying by plan, region, and rollout. This listing uses $20/month as the starting paid ChatGPT plan reference, while API billing is separate.
For GPT-5.6 API usage, OpenAI lists per-million-token prices of $5 input and $30 output for Sol, $2.50 input and $15 output for Terra, and $1 input and $6 output for Luna. Cached input receives the stated cached-input discount, while cache writes have separate billing rules for GPT-5.6 and later models. Teams should not treat API prices as ChatGPT subscription prices; they are separate billing surfaces.
How It Compares
Compared with earlier ChatGPT releases, GPT-5.6 makes model routing more explicit. The earlier GPT-5.4 mini and nano split focused on smaller, cheaper models. GPT-5.6 keeps the tiering idea but expands it into a broader family for professional work: Sol for flagship reasoning, Terra for balanced workloads, and Luna for low-cost scale.
Compared with using standalone developer tools, ChatGPT's advantage is breadth. A team can move from discussion to code, documents, spreadsheets, presentations, or API-backed workflows inside the OpenAI ecosystem. The tradeoff is complexity: plan access, effort settings, API price, and product surface all matter when choosing the right path.
Best For
- Individuals who want one assistant for writing, research, coding, and analysis.
- Product and engineering teams using ChatGPT and Codex for app building or code review.
- Operations and knowledge-work teams producing documents, presentations, spreadsheets, and research summaries.
- Developers building API workflows that need tool calling, cache control, and model-tier routing.
- Teams that want a single vendor ecosystem across chat, coding, and programmable AI.
FAQ
What is the current ChatGPT model family?
The current family in this listing is GPT-5.6, with Sol, Terra, and Luna tiers.
What is GPT-5.6 Sol?
Sol is the flagship GPT-5.6 tier for the hardest reasoning, coding, knowledge-work, science, and security tasks. Access depends on product surface and plan.
Is GPT-5.6 available in Codex?
Yes. OpenAI announced GPT-5.6 across ChatGPT, Codex, and the API. Codex users can choose supported GPT-5.6 tiers and effort settings depending on plan eligibility.
How much does GPT-5.6 cost in the API?
OpenAI lists GPT-5.6 API prices per million tokens: Sol at $5 input and $30 output, Terra at $2.50 input and $15 output, and Luna at $1 input and $6 output.
Is ChatGPT free?
ChatGPT has a free tier, but advanced model access, higher limits, team features, and the strongest GPT-5.6 settings require paid plans or API billing.
Who should use Terra or Luna instead of Sol?
Use Terra or Luna for tasks where lower latency or lower cost matters more than maximum reasoning depth. Routine classification, extraction, summarization, and high-volume API tasks often need cost testing before defaulting to Sol.
