Scoop: OpenAI built a credit bonus into ChatGPT’s experimental desktop onboarding
A hidden flow asks newcomers to let ChatGPT complete a real desktop task, then can add a remotely configured bonus to their balance.
By RuntimeWire Staff · Published
RUNTIMEWIRE INVESTIGATION — Scoop
Original reporting by RuntimeWire, based on reverse engineering.
Why it matters
OpenAI is preparing to use credits as an onboarding incentive, paying users to complete the permission-heavy step that turns ChatGPT from a chatbot into a desktop agent.
Reporting record
Finding
OpenAI is testing a ChatGPT desktop onboarding flow that asks new users to try a real starter task and offers some eligible accounts a remotely configured credit bonus for finishing setup.
How we verified
Methods: reverse engineering.
ChatGPT desktop version 26.803.81509 contains a complete, experiment-gated “conversational onboarding” system. The flow asks the user’s role, presents role-specific starter tasks, handles permissions and app connections, executes the selected task, and records its outcome. A separate remote configuration supplies a credit_amount; when that amount is greater than zero for a ChatGPT-authenticated account, the interface promises “Finish set up and get [X] credits,” labels the reward an “Onboarding bonus,” and says the credits will be “added to your balance.” Skipping displays language indicating the offer will be forfeited. Completing onboarding sends a request to /wham/onboarding/desktop/complete containing the selected role, whether onboarding was skipped, and whether the credit-reward warning was shown. The bundle also contains a dedicated error message for cases where onboarding credits cannot be processed.
RuntimeWire performed static analysis of an OpenAI desktop application archive dated August 11, 2026. We extracted its Electron ASAR package without executing the bundled code, verified the archive’s integrity, identified the product version, and traced the onboarding components across the compiled renderer modules. We followed the reward configuration from its remote-config lookup through the display conditions, skip dialog, completion state, error handling, and backend request. We separately inspected the starter-task definitions and their permission, connection, execution, completion, and restoration paths. Finally, we searched OpenAI’s current release notes, consumer credit documentation, desktop promotion terms, and the public web for the feature’s identifying language. No public documentation or prior coverage of this onboarding-completion reward was found.
Tested versions: ChatGPT/Codex desktop 26.803.81509 — package openai-codex-electron, build 6415.
Reproduction
RuntimeWire independently reproduced the core finding.
Client UI reproduced. RuntimeWire activated the dormant onboarding path in a copied desktop build and supplied a local display value to the server-configured credit field. The application rendered the complete reward dialog and starter-task interface. Account eligibility, live reward amounts, and balance crediting were not tested.
File hashes
sha256:39c855f6a0db1cad89dc629edf2f9473b78a25b64fcb1445ece1022af6527755 app.asar
Company response
RuntimeWire requested comment; the company had not responded by publication time.
OpenAI is testing a hands-on setup flow that can reward new ChatGPT desktop users with credits after the assistant completes a real task on their computer.
RuntimeWire found the flow in ChatGPT desktop build 26.803.81509, dated August 11th. The implementation is hidden behind experiment flags, so its presence does not establish that OpenAI is currently showing it to users. The code does contain the interface, task execution, permission handling, completion tracking and credit-award path needed to run the test.
The setup begins as a conversation rather than a standard product tour. ChatGPT asks newcomers what they do, records the selected role and offers to demonstrate its desktop capabilities through a task with a visible result.
The bundled examples include scheduling focus time, sending the user a message, creating a chart from a spreadsheet, leaving a note on the desktop and changing the computer's system theme. These choices move the user directly into the permissions and confirmation steps required for computer control, turning onboarding into an initial agent task.
The theme demonstration includes logic to restore the user's original system setting after completion. The build also handles failed tasks and incomplete runs, indicating that OpenAI has prepared for the flow to execute actions rather than play a fixed animation.
The reward is controlled remotely
The conversational setup and the credit reward sit behind separate experiment gates. OpenAI could run the task-based onboarding without offering credits, restrict the reward to a subset of eligible users or vary the bonus without shipping another desktop build.
Eligible interfaces include the text "Finish set up and get X credits." The completion screen labels the award an "Onboarding bonus" and tells the user that the credits were "added to your balance." Skipping the setup forfeits the displayed offer, according to the implementation.
The reward amount comes from remote configuration. Its bundled default is zero, which means the archived build cannot establish the size of any bonus. Plan, account and geographic eligibility can also be determined outside the desktop package.
Completion calls the endpoint /wham/onboarding/desktop/complete. The request records whether the user skipped onboarding, whether the reward warning was shown and which role the user selected. That gives OpenAI a way to measure whether the credit prompt changes completion rates and which user groups proceed through the demonstration.
The build contains the complete reward path, but experiment flags determine exposure. The archive alone cannot establish whether any users are seeing the offer as of August 15th.
Credits become an activation tool
OpenAI's public consumer credit documentation describes credits as a way to continue using supported features after a plan's included limits are exhausted. Users can purchase credits, and eligible Plus and Pro accounts can configure automatic top-ups.
OpenAI also maintains a separate ChatGPT desktop referral program that can award usage credits or other benefits after an invited user completes a qualifying action. The setup experiment removes the referral requirement and ties the potential reward directly to a new user's first successful desktop task.
The current ChatGPT release notes, updated through August 14th, do not describe an onboarding bonus or credit reward. OpenAI has documented other recent desktop changes, including the app's unified Chat, Work and Codex interface, without announcing this test.
The new desktop app gives ChatGPT access to local files, desktop applications and developer tools with the user's permission. That raises a harder activation problem than a conventional chatbot faces: users must understand what the agent can do, grant access and see a useful result before they trust it with larger tasks.
The proposed setup addresses that problem with a controlled first action. Each sample task is small, reversible or easy to inspect. The credit bonus adds an economic incentive to finish the permission flow, while the completion endpoint lets OpenAI measure whether the incentive produces actual setup rather than a dismissed tutorial.
If OpenAI enables the experiment, credits would serve two purposes at once. They would subsidize a newcomer's initial agent usage and push that user through the moment when ChatGPT crosses from answering questions to operating software on the desktop.