Last updated: July 7, 2026
GPT-5.6 Terra is the balanced model in OpenAI’s GPT-5.6 family. If GPT-5.6 Sol is the flagship for the hardest work and GPT-5.6 Luna is the fast low-cost tier, Terra is the middle option: strong enough for serious professional tasks, cheaper than Sol, and more capable than a pure budget model.
Quick answer: GPT-5.6 Terra’s preview model ID is gpt-5.6-terra. OpenAI lists preview pricing at $2.50 per 1M input tokens and $15 per 1M output tokens. During the preview, GPT-5.6 models are available only to selected trusted partners and organizations through the OpenAI API and/or Codex. OpenAI says GPT-5.6 is not available in ChatGPT during the preview.
Verified July 7, 2026
- OpenAI launch post: Previewing GPT-5.6 Sol
- OpenAI Help Center: A preview of GPT-5.6 Sol, Terra, and Luna
- Safety details: GPT-5.6 Preview System Card
- Model catalog status: OpenAI API models page
GPT-5.6 Terra quick facts
| Detail | GPT-5.6 Terra |
|---|---|
| Model family | GPT-5.6 |
| Model ID | gpt-5.6-terra |
| Role | Balanced lower-cost option |
| Preview input price | $2.50 / 1M tokens |
| Preview output price | $15.00 / 1M tokens |
| Cache write | $3.125 / 1M tokens |
| Cached input read | $0.25 / 1M tokens |
| Preview access | Selected API/Codex partners and organizations |
| ChatGPT access | Not included during preview |
| Best for | Everyday professional work, analysis, coding help, routing defaults |
| Escalate to Sol when | Failure is expensive or the workflow is deeply agentic |
| Downgrade to Luna when | The task is simple, high-volume, and easy to verify |
What is GPT-5.6 Terra?
GPT-5.6 Terra is the balanced tier in the GPT-5.6 lineup. OpenAI describes the family as:
- Sol — flagship and most capable;
- Terra — strong lower-cost option;
- Luna — fastest and most cost-efficient option.
Terra is the model that should interest most teams building a routing strategy. Sol may be too expensive for every request. Luna may be too small for nuanced professional work. Terra sits in the middle: use it when you need quality, but not necessarily the maximum reasoning budget.
GPT-5.6 Terra pricing
OpenAI lists Terra at:
| Token type | Price per 1M tokens |
|---|---|
| Input | $2.50 |
| Output | $15.00 |
| Cache write | $3.125 |
| Cached input read | $0.25 |
Cache write is calculated from OpenAI’s GPT-5.6 rule: 1.25x the uncached input rate. Cached reads receive a 90% discount, making repeated long prompts much cheaper when caching applies.
For example, if your workflow repeatedly includes the same policy document, codebase summary, or product manual, Terra may be a good default because it balances answer quality with cache economics.
Terra vs Sol vs Luna
| Model | Best fit | Main tradeoff |
|---|---|---|
| GPT-5.6 Sol | Hardest reasoning, agentic coding, security, science, long workflows | Highest cost and likely more latency |
| GPT-5.6 Terra | Balanced professional work, coding help, analysis, document workflows | Not as strong as Sol for the hardest tail |
| GPT-5.6 Luna | Fast high-volume tasks, extraction, classification, cheap drafts | Not the first choice for complex tasks |
A practical routing strategy:
- Start with Luna for simple extraction, tagging, and first drafts.
- Route normal work to Terra.
- Escalate to Sol when Terra fails, the task is high-value, or the workflow needs deeper reasoning.
Best GPT-5.6 Terra use cases
Use Terra for tasks such as:
- business writing that needs structure but not flagship-level reasoning;
- policy and document summarization;
- code explanation and routine debugging;
- test generation and code review drafts;
- research synthesis from curated sources;
- spreadsheet and report analysis;
- customer-support answer drafting with human review;
- internal workflow automation;
- planning documents, briefs, and SOP drafts.
Terra is especially useful when your system has many tasks that are too hard for a cheap model but not valuable enough to send to Sol by default.
When not to use Terra
Do not default to Terra when:
- a lower-cost model already solves the task reliably;
- the work is mission-critical and Sol’s stronger reasoning is justified;
- the task requires broad ChatGPT access during the preview;
- you cannot verify the output;
- you have not confirmed your organization has preview access.
Terra and safety
OpenAI’s GPT-5.6 materials emphasize stronger safeguards across Sol, Terra, and Luna, with configurations matched to model capability. The system card says the larger models generally do better on complex tasks, while also discussing high capability in cybersecurity for the GPT-5.6 family.
For teams, the takeaway is not “Terra is unsafe” or “Terra is automatically safe.” The takeaway is that Terra should be evaluated with the same controls as any model used for sensitive work:
- clear allowed and disallowed use cases;
- human review for regulated outputs;
- logging and monitoring;
- prompt and tool boundaries;
- extra caution for cyber, biology, health, legal, and financial workflows.
Should you test GPT-5.6 Terra?
Test Terra if you have approved preview access and want a practical default model for real workloads.
- Everyday professional workflows
- Analysis and summarization
- Routine coding help
- Balanced cost/quality routing
- Security or biology workflows
- Customer-facing automation
- Long autonomous tasks
- Simple high-volume tagging that Luna handles
- High-stakes work that should escalate to Sol
- Consumer ChatGPT usage during preview
Bottom line
GPT-5.6 Terra is likely the most practical GPT-5.6 tier for many teams once they have access. It is cheaper than Sol, stronger-positioned than Luna, and designed for the middle of the workload distribution.
The smart move is to test Terra against your real tasks, then route:
- Luna for simple high-volume work;
- Terra for normal professional work;
- Sol for hard, expensive-to-fail tasks.
FAQ
What is GPT-5.6 Terra?
What is the GPT-5.6 Terra model ID?
gpt-5.6-terra.