Standard Compute

OpenAI · updated 2026-09-05

Run GPT-6 Astra: providers, pricing and flat-rate access

OpenAI’s September 2026 frontier model for software engineering, reasoning, and multi-step professional work.

Run GPT-6 Astra through the OpenAI API, through OpenRouter, or with Standard Compute’s monthly compute plans. On Standard Compute, select it in the dashboard and keep your agent configured with the standardcompute model. Your monthly budget covers usage; a fixed plan price does not mean unlimited tokens.

Your options, honestly

Model documentation: OpenAI: model documentation

OpenAI APIPay per tokenDirect API access. Check the vendor’s current pricing, account access, and usage limits.
OpenRouterPay per tokenAccess through an OpenAI-compatible API. Provider availability and rates can vary.
Standard ComputeFrom $19/monthAvailable in the model picker. Usage draws from your monthly compute budget, with no automatic overage charges.
the flat-rate trade-off

A direct API can be a better fit for light or occasional usage. A monthly compute plan is useful when you want a predictable bill across several models. Premium models consume the budget faster; compare the cost of completed tasks, including retries and cached input.

Running it in your agent

In an OpenAI-compatible agent, set the base URL to https://api.stdcmpt.com/v1, paste your API key, and use standardcompute as the model. Select GPT-6 Astra in Dashboard → Models. Select only this model to pin requests to it, or add up to four others for smart routing.

FAQ

Is GPT-6 Astra included in Standard Compute?
Yes. Select GPT-6 Astra in your dashboard. Availability is shared across plans; plans differ by compute budget and scheduling. Adding it to the catalog does not replace your saved selection.
Does a monthly plan give unlimited GPT-6 Astra tokens?
No. Every plan includes a defined compute budget. Requests stop when the budget runs out, or optional pacing can spread usage across the month. There are no automatic overage charges.

More: how flat-rate LLM APIs work · per-agent guides for OpenCode, Claude Code and Cline · model comparisons