GPT-5.6 Terra
GPT-5.6 Terra is a large language model made by OpenAI, added to the catalog on 9 Jul 2026. It has a context window of 1,050,000 tokens and costs $2.00 per 1M input tokens and $12.00 per 1M output tokens. It accepts text, images and files and produces text. Typical uses: everyday coding tasks, reasoning over long documents and reviewing supplied images and files.
Price checked ยท source: OpenRouter
What GPT-5.6 Terra is
OpenAI GPT-5.6 Terra is a model for text-based work across coding, reasoning, and agentic tasks. It accepts text, images, and files, and produces text. Its context window holds up to 1,050,000 tokens, while its maximum output is 128,000 tokens. That gives it room to work with long inputs and return lengthy text, within those limits. The catalog lists it as a balanced model in the GPT-5.6 series; it does not provide further detail about particular coding languages, image types, or file formats.
Terra supports reasoning, tool use, structured outputs, and response-format controls. These capabilities suit tasks that combine supplied material with analysis or need a response shaped to a defined format. With text, image, and file inputs, it can be considered for reviewing documents or visual material and producing written explanations, summaries, or answers. Its large context can also suit bringing multiple lengthy source materials into one request, while the output limit allows substantial responses. This combination can support handling a long prompt that includes written instructions, files, and images, then returning a text response. Tool support provides an option for workflows that use tools, while structured-output and response-format support accommodate specified output shapes. The model's catalog description identifies everyday coding, reasoning, and agentic work as intended areas, without detailing specific workflows or guarantees.
Key features
- Accepts text, image, and file inputs and produces text
- Context window of up to 1,050,000 tokens
- Maximum output of 128,000 tokens
- Supports tools, reasoning, structured outputs, and response formats
Typical uses
- Everyday coding tasks
- Reasoning over long documents
- Reviewing supplied images and files
- Generating structured text responses
- Tool-supported agentic workflows
GPT-5.6 Terra pricing
US dollars, list price through OpenRouter.
| Item | Standard | Batch |
|---|---|---|
| Input, per 1M tokens | $2.00 | $1.00 |
| Output, per 1M tokens | $12.00 | $6.00 |
| Cached input read, per 1M tokens | $0.20 | $0.10 |
| Cache write, per 1M tokens | $2.50 | n/a |
| Web search, per search | $0.01 | $0.01 |
| Input above 272,000 prompt tokens, per 1M | $4.00 | $2.00 |
| Output above 272,000 prompt tokens, per 1M | $18.00 | $9.00 |
| Variant | Model id |
|---|---|
| Batch | openai/gpt-5.6-terra:batch |
A variant is the same model under another id, with prices or limits of its own. Its prices are the columns of the list above.
No price change since it was added on 9 Jul 2026.
| What a task costs | Cost |
|---|---|
| Answer a short question500 in / 300 out | $0.0046 |
| Summarise a 50-page PDF25,000 in / 800 out | $0.060 |
| Review a 2,000-line source file30,000 in / 2,000 out | $0.084 |
| Read a 300-page contract set150,000 in / 1,500 out | $0.318 |
GPT-5.6 Terra specifications
| Model id | openai/gpt-5.6-terra |
|---|---|
| Version | gpt-5.6-terra-20260709 |
| Context window | 1,050,000 tokensAbout 787,500 English words |
| Max output | 128,000 tokens |
| Knowledge cutoff | 2026-02-16 |
| Added to catalog | 9 Jul 2026 |
| Input types | file, image, text |
|---|---|
| Output types | text |
| Tool calling | Yes |
| Structured outputs | Yes |
| Reasoning effort control | Yes |
| Content moderation | Applied by provider |
Sample outputs
The same prompts are given to every model, and the unedited output is shown with the date it was produced. Upend does not publish leaderboard scores it did not run.
No output has been recorded for GPT-5.6 Terra yet.
Compare GPT-5.6 Terra
Side by side with other models: prices, context window, what each accepts and supports, and the same four tasks priced on both.