Service eligibility and regional restrictions
PuppyIP serves only compliant overseas businesses and their authorized personnel. Proxy services are not available in mainland China. The service may only be used for lawful business activities outside mainland China. Use of this service within mainland China is prohibited.
Hosting a proxy IP or server overseas does not change these restrictions. The service must not be provided to end users in mainland China through relaying, forwarding, sharing or resale. Before use, read the Terms of Service.
Key Takeaways
- OpenAI announced GPT-6.1 Terra on October 26, 2026. The announcement lists Plus, Pro, Enterprise and Edu access in Work and Codex.
- The API model ID is `gpt-6.1-terra`; Standard pricing is $1 per million input tokens, $0.05 per million cached input tokens, from $1.25 per million cache-write tokens and $5 per million output tokens.
- The current API docs list a 1,050,000-token context window and 128,000-token maximum output. Long-context pricing applies above 272K input tokens.
- The model accepts text and image input, but not audio or video. Tool calling requires the Responses API.
What is GPT-6.1 Terra?
OpenAI describes GPT-6.1 Terra as a fast, efficient model focused on visual reasoning, precise computer use and natural writing. Its examples include matching product photos to catalog entries, interpreting technical drawings and carrying out work in application screens.
These are OpenAI’s stated use cases and evaluation results, not a guarantee that every workflow will succeed. For tasks that act on a live interface, validate the model with your own tools, permissions and data boundaries first.
Who can use it in ChatGPT Work and Codex?
OpenAI’s October 26, 2026 announcement says GPT-6.1 Terra is available starting that day to Plus, Pro, Enterprise and Edu users. The named product surfaces are ChatGPT Work and Codex. The announcement does not say Terra is available in every standard ChatGPT conversation or model picker, and it does not name Business in that eligibility statement.
Access may also depend on workspace controls and rollout. Check the model picker in your own Work or Codex account; the announcement alone does not prove that a particular account already has the option.
How do you call the API, and what are its limits?
The API model ID is `gpt-6.1-terra`. OpenAI’s current model page lists a 1,050,000-token context window and a maximum output of 128,000 tokens. Those are model limits; they do not mean every request should use the full context.
The page lists text and image input, with audio and video unsupported. `reasoning.effort` accepts `none`, `low`, `medium`, `high`, `xhigh` and `max`. OpenAI's API changelog lists GPT-6.1 Terra for `v1/responses` and `v1/chat/completions`; the model page lists tool support through the Responses API, so do not assume every built-in tool is available through Chat Completions. API access, account rate limits and ChatGPT subscription eligibility are separate questions.
How is API usage priced?
Current Standard API rates are $1 per million input tokens, $0.05 per million cached input tokens and $5 per million output tokens. The model page also lists cache writes at $1.25 per million tokens for 30-minute retention and $2.50 for 12-hour retention; it says these cache-write prices are before long-context adjustments. OpenAI says the Standard rates are 50% lower than the Standard API prices for GPT-5.6 Terra and GPT-6.1 Sol.
For a text-only estimate, 100,000 uncached input tokens plus 20,000 output tokens comes to about $0.20 at those Standard rates, assuming the request stays below the long-context threshold and has no cache writes or tool fees. Actual charges may also include separate tool fees; Batch or Flex, Fast mode and regional processing can change the price.
For a request with more than 272K input tokens, the model page says long-context pricing applies to the full request: input and cached-input rates are 2× Standard, and output is 1.5× Standard. A larger context window does not mean every request is billed at the short-context rate.
When is Terra worth trying?
If your workload involves repeated image understanding, scoped computer use or natural writing, Terra’s positioning and lower Standard rates make it worth testing against your current model on the same task sample. OpenAI also says GPT-6.1 Sol or GPT-6 Astra may be better choices for its most complex software-engineering tasks; your own code, tools and acceptance criteria determine the result.
Before switching, compare the same input, model settings and tool permissions, then record task outcomes and total cost. A Work or Codex entitlement is not API credit, and computer-use support does not override an application’s permissions, account rules or human-confirmation requirements.
Sources
Frequently Asked Questions
Is GPT-6.1 Terra available in regular ChatGPT chats?
OpenAI’s announcement specifically names ChatGPT Work and Codex. It does not confirm that Terra is available in every standard ChatGPT conversation or model picker, so check your own account.
Does a ChatGPT Work or Codex plan include API usage?
Do not assume that it does. The announcement describes product access and API access separately. API usage has its own token pricing and account usage tiers.
Are all 1,050,000 context tokens charged at Standard rates?
No. The API docs say long-context pricing applies to the entire request when input exceeds 272K tokens, changing input, cached-input and output rates.
Can Terra process audio or video directly?
OpenAI’s current model page lists text and image input and says audio and video are unsupported. Check the model page again before integrating because capabilities can change.