openai-fast

A pi extension that enables OpenAI Codex Fast mode for ChatGPT-auth GPT-5.4, GPT-5.5, and GPT-5.6 Codex variants (gpt-5.6-sol, gpt-5.6-terra, gpt-5.6-luna).

When active, the extension injects this into eligible OpenAI Codex request payloads:

{
  "service_tier": "priority"
}

The user-facing feature is OpenAI Codex Fast mode. The wire value is priority because current Codex clients map Fast mode to the OpenAI priority service tier.

Eligibility

Fast mode is only injected when all of these are true:

• The current provider is openai-codex.
• The current API is openai-codex-responses.
• The current model is gpt-5.4, gpt-5.5, gpt-5.6-sol, gpt-5.6-terra, or gpt-5.6-luna.
• The provider is using ChatGPT OAuth/subscription auth, not API-key auth.
• The request payload does not already include service_tier.

Commands

/fast

Run /fast to toggle Fast mode on or off for the current session/runtime. The command reports the new state in chat, and the footer shows fast while Fast mode is active for an eligible model.

The extension defaults to off so installing the full collection does not accidentally spend Fast-mode credits.

Config

Optional global config:

~/<pi-config-dir>/agent/extensions/openai-fast.json

Optional project config:

<project>/<pi-config-dir>/openai-fast.json

Here <pi-config-dir> is Pi's runtime config directory name (CONFIG_DIR_NAME; .pi by default). Project config overrides global config after Pi reports that the project is trusted.

{
  "enabled": false,
  "showStatus": true
}

• enabled: default Fast-mode state when there is no session override.
• showStatus: show a compact fast status when Fast mode is active for the current model.

Install

Standalone npm package

pi install npm:@diegopetrucci/pi-openai-fast

Collection package

pi install npm:@diegopetrucci/pi-extensions

GitHub package

pi install git:github.com/diegopetrucci/pi-extensions

Then reload pi:

/reload

Notes

• This extension intentionally does not affect API-key OpenAI models.
• Pi may only account Fast-mode cost correctly when the backend reports service_tier: "priority" in the streamed response. The extension does not patch usage totals to avoid double-counting.
• If pi adds first-class service-tier support later, this extension skips payloads that already contain service_tier.