Standard Compute

Google · updated 2026-09-05

Run Gemini 3.8 Flash: providers, pricing and flat-rate access

Google’s latest Flash model for fast agent workflows, with a one-million-token context window.

Run Gemini 3.8 Flash through the Google 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: Google: model documentation

Google 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 Gemini 3.8 Flash in Dashboard → Models. Select only this model to pin requests to it, or add up to four others for smart routing.

FAQ

Is Gemini 3.8 Flash included in Standard Compute?
Yes. Select Gemini 3.8 Flash 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 Gemini 3.8 Flash 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