OpenAI Platform Leaks AI App Hosting Plans: Free, Prototype, and Accelerate

Leaked onboarding flows on the OpenAI Platform reveal three upcoming tiers for building and hosting AI apps: Free, Prototype, and Accelerate, ahead of DevDay.

tau · September 23, 2026

#OpenAI #DevDay #OpenAIPlatform #AIAppHosting #Vercel

OpenAI Platform Leaks AI App Hosting Plans: Free, Prototype, and Accelerate

On September 23, 2026, details of a new onboarding workflow on the OpenAI Platform surfaced, revealing three upcoming subscription plans designed to let users build and host their own AI applications: Free, Prototype, and Accelerate. The leak was originally reported by TestingCatalog alongside a demonstration screencast capturing the unreleased interface.

Onboarding interface on OpenAI Platform displaying Free, Prototype, and Accelerate subscription plans for AI apps Image source: @testingcatalog

The newly exposed flow indicates an architectural shift from standard API token issuance toward an integrated application building and hosting environment. Widely expected to be officially unveiled at the upcoming OpenAI DevDay conference, this move signals OpenAI's potential entry into the application deployment and hosting ecosystem currently led by platforms like Vercel and Lovable.

Leaked Onboarding Flow and the Three-Tier Structure

According to the walkthrough video published by TestingCatalog, the onboarding interface guides developers through a tiered selection process designed to help users build and host their own AI applications.

The leaked interface presents three distinct tiers: 'Free', 'Prototype', and 'Accelerate'. Unlike standard consumer ChatGPT subscriptions, these plans reside within the developer platform, functioning as a subscription-style onboarding flow tailored specifically around application development and hosting.

  • Free Tier: An entry-level tier designed to let developers start building and experimenting with AI applications on the platform at no initial cost.
  • Prototype Tier: Positioned as an intermediate tier between Free and Accelerate, targeted at supporting early-stage prototyping and small-scale application builds.
  • Accelerate Tier: A production-oriented tier starting at $50 ('Starting at $50'), providing full access to OpenAI's latest models, higher usage and rate limits for larger production apps, and accelerated building with the API in Codex and Playgrounds.

In the initial leak, specific starting prices ("Starting at $50") and concrete feature bullets were confirmed primarily for the Accelerate tier. Detailed quota specifications and official pricing for the Free and Prototype tiers remain to be formally detailed upon official announcement.

Accelerate Tier Specifications: $50 Entry, Production Limits, and Full Model Access

Highlighted as the flagship production tier within the leaked flow, the 'Accelerate' plan is explicitly marked as "Starting at $50". The core features outlined in the UI include:

  • Access all of our latest models: Comprehensive access to the full spectrum of OpenAI foundation models.
  • Higher usage & rate limits to power larger, production apps: Significantly expanded API usage and rate limits engineered to support larger, production-scale applications.
  • Accelerate building with the API in Codex and Playgrounds: Accelerated developer tooling and API capabilities inside Codex and interactive Playgrounds.

The "Starting at $50" pricing tier represents an accessible entry point for startups and independent developers seeking production stability, combining increased rate caps with developer environment acceleration.

DevDay Announcement Outlook: Platform Evolution and Caveats

TestingCatalog projects that these subscription tiers are slated for official unveiling during the upcoming OpenAI DevDay.

If OpenAI directly integrates hosted infrastructure alongside model APIs, it would place the company into closer competition with developer hosting and scaffolding services such as Vercel and Lovable—a prospect TestingCatalog specifically highlighted as plausible industry speculation. This direction also aligns with recent commentary from OpenAI CEO Sam Altman, who remarked: "And then we want you all to come up with great ideas and build them and to get to be happy users. The best ideas will come from you all."

However, because this intelligence stems from pre-release interface leaks, several operational caveats remain:

  • Policy Adjustments Prior to Release: Plan names, exact pricing structures, model allotments, and billing mechanics may be altered prior to formal launch at DevDay.
  • Hosting Function Scope: The present report relies on onboarding UI copy and demonstration captures; detailed architectural specifications regarding actual code hosting and infrastructure runtime limits await official developer documentation.

Sources