How to Use GPT-6.1 Terra

To use GPT-6.1 Terra, select the model inside ChatGPT Work or Codex on an eligible paid subscription, or send requests to the gpt-6.1-terra model identifier in the OpenAI application programming interface (API). At Layer3Labs, we build document and workflow automation for businesses.

OpenAI launched GPT-6.1 Terra on October 26, 2026. It is a fast, efficient model combining detailed visual reasoning and precise computer use with clearer, more natural writing, at half the standard API price of GPT-6.1 Sol.

How to Turn On GPT-6.1 Terra

ChatGPT Work and Codex require eligible account access. For the API, configure an API key and billing, then set the model to gpt-6.1-terra using the OpenAI quickstart.

Follow these steps to access GPT-6.1 Terra on your chosen interface.

• 1. ChatGPT Work: Confirm that your organization maintains an active Plus, Pro, Business, Enterprise, or Edu subscription. Open ChatGPT Work and choose GPT-6.1 Terra from the Advanced model menu.
• 2. Codex: On the same Plus, Pro, Business, Enterprise, or Edu account tiers, open Codex and select GPT-6.1 Terra from the Advanced model menu as the active Codex model for your coding environment. For the command line, use Codex CLI 0.159.0 or later and run codex -m gpt-6.1-terra, as listed in the model instructions.
• 3. OpenAI API: In your application code, pass the model parameter gpt-6.1-terra using Chat Completions, Responses, or Batch endpoints. When your integration requires tool calling, configure your client to use the Responses API specifically.

Account Plans That Include Access to GPT-6.1 Terra

OpenAI's October 26, 2026 changelog lists GPT-6.1 Terra as available within ChatGPT Work and Codex for Plus, Pro, Business, Enterprise, and Education plans. ChatGPT Free and Go tiers do not include GPT-6.1 Terra, and developer API access is not part of workspace seat subscriptions and is billed separately under API rates.

In ChatGPT Work and Codex, OpenAI reports that your account usage allowance covers more tasks on GPT-6.1 Terra than on GPT-5.6 Terra, and substantially more than on GPT-6.1 Sol or GPT-6 Astra.

In the API, you pay per token. Standard developer API rates for gpt-6.1-terra run at $1.00 per million input tokens, $0.05 per million cached input tokens, and $5.00 per million output tokens on inputs up to 272,000 tokens. That is half of GPT-6.1 Sol's $2.00 input and $10.00 output rates.

Calling the gpt-6.1-terra Model Through the API

Developer integration for gpt-6.1-terra operates across the Chat Completions API, the Responses API, and the Batch API. The model supports text and image inputs while generating text outputs. It does not accept audio or video payloads. Its context window spans 1,050,000 tokens, with a maximum allowable input payload of 922,000 tokens and a maximum output ceiling of 128,000 tokens. Its knowledge cutoff date is June 1, 2026.

When issuing requests, developers can control inference depth through the reasoning effort parameter. Accepted values are none, low, medium, high, xhigh, and max, with medium serving as the default setting. Reasoning tokens grow with effort level.

Tool execution requires the Responses API. The model supports web search, file search, image generation, code interpreter, hosted shell, apply patch, skills, computer use, Model Context Protocol (MCP), and tool search, so GPT-6.1 Terra can operate desktop apps, search, and run code through the Responses API. For rate limits on the standard tier, Build accounts receive 5,000 requests per minute (RPM) and 1,000,000 tokens per minute (TPM). Launch accounts receive 10,000 RPM and 4,000,000 TPM. Grow accounts scale to 15,000 RPM and 40,000,000 TPM.

Pay attention to the 272,000 token pricing step. Input payloads that stay at or below 272,000 tokens bill at $1.00 per million input tokens and $0.05 per million cached input tokens. Once total input size crosses 272,000 tokens, input rates double to $2.00 per million tokens, cached inputs increase to $0.10 per million tokens, and output rates shift from $5.00 to $7.50 per million tokens. OpenAI lists cache writes at $1.25 per million tokens for 30-minute retention and $2.50 for 12-hour retention in its API pricing.

Operational Use Cases to Deploy on GPT-6.1 Terra First

OpenAI highlights several target jobs for GPT-6.1 Terra, including drafting proposals, interpreting technical drawings, filling out forms, editing videos, and reading incoming invoices to log details into accounting software.

In our internal cost analysis, we priced 1,000 incoming invoices at the standard per-token rates OpenAI and Anthropic publish. In this hypothetical scenario, each invoice requires 3,000 input tokens and generates 500 output tokens. On GPT-6.1 Terra, standard API processing costs $5.50. On GPT-5.6 Terra, that same run costs $12.00. Running the workload against GPT-6.1 Sol costs $11.00, Claude Sonnet 5.5 costs $11.00, and GPT-6 Astra costs $55.00. The lightweight GPT-6 Luna costs $0.55. These modeled figures illustrate token price differences using assumed prompt lengths rather than measured real-world application logs.

We modeled two additional hypothetical scenarios to evaluate volume behavior. For customer support drafts handling 10,000 replies at 2,000 input tokens (with 60 percent stored in cached context) and 400 output tokens, GPT-6.1 Terra totals $28.60, compared to $57.20 on Sol and $292.00 on Astra. For a review of 100 long documents using 200,000 input tokens and 4,000 output tokens each, Terra costs $22.00, while Sol costs $44.00 and Astra reaches $220.00.

These token-only estimates assume the stated cache reads hit. They exclude cache writes, cache warmup, tool fees, retries, and taxes. Failed cache reads increase cost.

Token rates alone do not dictate overall job costs. If an application requires high reasoning effort, Terra generates more reasoning tokens to solve the prompt. If GPT-6.1 Terra produces 3.2 times as many output tokens as GPT-6.1 Sol for the same invoice, the cost advantage disappears. Run a sample batch on both models and compare output tokens per invoice before you commit.

When to Switch to GPT-6.1 Sol or GPT-6 Astra

OpenAI explicitly notes that for the most complex coding tasks, GPT-6.1 Sol and GPT-6 Astra remain the better choice.

Artificial Analysis's independent index shows the same gap. On the Coding Agent Index v2.0 published by Artificial Analysis, GPT-6.1 Terra scores 48, trailing GPT-6.1 Sol at 54 while outperforming GPT-6 Luna at 31. On command-line coding evaluations using Terminal-Bench 4.0, OpenAI reports that Terra achieves a 50.3 percent solve rate at maximum reasoning effort, improving on GPT-5.6 Terra at 21.5 percent. Anthropic reports 70.6 percent for Claude Sonnet 5.5 on Terminal-Bench 4.0 from its own test setup, so the two scores are not a head-to-head result. The GPT-6.1 Terra vs Claude Sonnet 5.5 comparison covers that choice.

If your hardest work is complex coding, start on GPT-6.1 Sol and keep Terra for the office and computer-use jobs above.

Troubleshooting Common GPT-6.1 Terra Setup Problems

Each of these setup problems has a direct fix.

OpenAI's October 26, 2026 changelog lists GPT-6.1 Terra as available within ChatGPT Work and Codex for Plus, Pro, Business, Enterprise, and Education plans. ChatGPT Free and Go tiers do not include GPT-6.1 Terra, and developer API access is not part of workspace seat subscriptions and is billed separately under API rates.

Tool calls may fail when sending requests through standard Chat Completions. OpenAI requires developers to use the Responses API for tool execution with gpt-6.1-terra. If your application invokes web search, file search, hosted shells, or computer use, update your API client endpoints to target the Responses API instead of Chat Completions.

Costs can jump when prompt sizes exceed 272,000 tokens. OpenAI applies a higher pricing tier to long-context inputs. Prompts above 272,000 tokens double the base input fee to $2.00 per million tokens and raise output fees to $7.50 per million tokens. To stay under the step, trim or split long inputs. If a prompt must run longer, cache the repeated context, because above 272,000 tokens cached input bills at $0.10 per million against $2.00 uncached.

How to Use GPT-6.1 Terra

To use GPT-6.1 Terra effectively, start by choosing an interface that exposes the exact model you need for your workload. Test a new configuration on a single representative task, such as drafting a routine email or parsing a short document, before expanding to full automation. Inspect the saved output directly to confirm formatting, accuracy, and adherence to your prompt instructions.

For coding tasks, Cursor serves as an editor option when the target model appears on its supported list. Check current model availability before setup to confirm the exact model is listed. Once connected, save your prompts and review the generated diffs locally before applying changes to your main codebase.