How do I use GPT-5.6 Terra in Codex without per-token billing?
Run Codex through Standard Compute with the model ID openai/gpt-5.6-terra: GPT-5.6 Terra is on every monthly plan from $19, each plan is a fixed compute budget, and at 100% new requests get HTTP 402 instead of a per-token bill. The same key works in OpenCode, Cline and Kilo Code, and no OpenAI usage tier is involved.
| Model | The middle tier of OpenAI's GPT-5.6 family, released July 9, 2026 for work that balances intelligence and cost. It has a 1.05M-token context window and 128K max output. |
|---|---|
| Model ID | openai/gpt-5.6-terraShort form gpt-5.6-terra also works |
| Base URL | https://api.stdcmpt.com/v1 |
| Model ID, EU pin | openai/gpt-5.6-terra:euUS pin: openai/gpt-5.6-terra:us, text only; the base ID and EU pin also take images and files |
| Plans | Every plan, from Starter at $19 a month with a $20 compute budget |
| Reasoning | minimal, low, medium or high, per request or as a Models page default |
Copy a working config
Codex is OpenAI's own agent and speaks the Responses API natively; the launcher passes the model straight through. To keep Terra for the harder steps only, select it with GPT-6 Luna on the Models page and set the model to standardcompute.
# Install Codex and the Standard Compute launcher once
npm install -g @openai/codex
curl -fsSL https://standardcompute.com/cli/install.sh | sh
# Start Codex on your plan with openai/gpt-5.6-terra
standardcompute launch codex -- --model openai/gpt-5.6-terra
Replace YOUR_API_KEY with a key from the dashboard. This config reads your key from STANDARDCOMPUTE_API_KEY; export it with a key from the dashboard. Full setup: Codex CLI, OpenCode, Kilo Code, Cline, Aider and other tools.
What it costs
| Per 1M tokens | Prompt up to 272K | Prompt over 272K |
|---|---|---|
| Input | $2.00 | $4.00 |
| Cached input | $0.20 | $0.40 |
| Cache write | $2.50 | $5.00 |
| Output | $12.00 | $18.00 |
OpenAI standard API rates. The higher column applies to the whole request once the prompt passes 272K tokens, and regional processing adds 10%. Source: OpenAI API pricing, checked October 13, 2026.
Average coding-agent turn on our gateway, Sep 13 to Oct 12, 2026. OpenAI list prices.
On this gateway Terra is the mid-priced OpenAI model, between GPT-6 Luna and GPT-5.6 Sol. On OpenAI's own API, GPT-6.1 Sol now lists at $2 / $10 per million tokens with cached input at $0.10, below Terra on output and cache hits; it is not on Standard Compute yet.
What works differently here
- Terra's 42 on the Artificial Analysis Intelligence Index was measured at OpenAI's max reasoning setting. Here Terra runs at minimal to high; xhigh and max are not offered.
- In Codex, built-in web search is off on this connection and
previous_response_idreturns 400. Codex sends the full conversation each turn, so it works as shipped.
Other ways to run GPT-5.6 Terra
| OpenAI API$2 / $12 per 1M | First-party, metered per token, with rate limits tied to your organization's usage tier. |
|---|---|
| ChatGPT plans with CodexSubscription | Codex inside each plan's usage limits, without a general API key for other agents. |
| Azure / OpenRouterPay per token | The same model under Azure billing or one OpenRouter account. |
| Standard ComputeMonthly plan | Terra with EU and US pins; Sol and both Lunas share the same key. |
FAQ
Is GPT-5.6 Terra on the $19 Starter plan?
Yes. Every plan carries the same models, Terra included. Plans differ in compute budget and request limits; Starter is $19 a month for a $20 budget.
GPT-5.6 Terra or GPT-5.4 for Codex?
GPT-5.4 is not served on Standard Compute. On OpenAI's list, Terra is also the lower price: $2 / $12 per million tokens against $2.50 / $15 for GPT-5.4.
GPT-5.6 Terra or GPT-6 Luna?
Luna lists at a twentieth of Terra's input price ($0.10 against $2) and Artificial Analysis scores it 38 to Terra's 42. Pin Luna for routine edits and tool calls; with both selected, routing sends the hard requests to Terra.
Can I keep GPT-5.6 Terra requests in the EU?
Yes. Use openai/gpt-5.6-terra:eu, which runs on OpenAI's EU regional processing. OpenAI lists regional processing at 10% above its standard rates.
Sources: OpenAI API pricing; OpenAI: GPT-5.6 Terra model page; Artificial Analysis: GPT-5.6 Terra
Plans from $19/mo
One key for GPT-5.6 Terra and the rest of the catalog, with a monthly compute budget you can watch in the dashboard and no overage charges.