Download Fame Codex · Report an issue · Leonelfame
Get started · What’s new · Architecture · Troubleshooting
Fame Codex is a desktop app maintained by Leonelfame. It connects your signed-in ChatGPT browser session to Codex through MCP, so Codex can edit project files and run commands with the tools and approvals for the current task.
Setup, the embedded ChatGPT browser, MCP, Activity, and Settings stay in one window. A separate DEV profile lets you test changes without replacing your normal account session.
Choose a Fame Codex model ending in (Web) from Codex's model picker. Available models and usage limits depend on your ChatGPT account.
The screenshot shows the local development build. Downloadable packages and their supported platforms are listed on the Releases page.
Open Releases and download an installer attached to a published release. Read its release notes for known issues and platform support. A version in this source tree does not mean its installer has been published.
Available models: Free/Go → Luna / Think. Paid accounts → GPT-6 Sol and GPT-5.6 Sol, with Instant–High, plus Extra High and Pro when available. Each version has its own entry; existing GPT-5.6 chats keep GPT-5.6. Availability depends on your account and workspace.
- Install the launcher from the Releases page.
- Sign in to ChatGPT in the embedded browser and run the browser smoke test.
- Install models and restart Codex once. In automatic mode, choose a model ending in (Web). Pro versions have separate entries; Sol reasoning is selected through Effort. Zero Risk keeps its dedicated entry.
- For coding with tools, open MCP in the launcher and complete the Full harness setup below.
The app includes its browser and runtime for ChatGPT. No separate Chrome, Node, or Bun installation is needed for that path. The optional TH AI Pass integration requires Chrome or Edge with the bundled browser extension.
TH AI Pass connects models available in your signed-in TH AI Pass account to Codex through a browser extension and a local bridge. It is disabled by default.
- Install Fame Codex on this computer. Open Settings → TH AI Pass Experimental in the launcher and enable TH AI Pass.
- Click Open extension folder.
- In Chrome, open
chrome://extensionsand turn on Developer mode. - Click Load unpacked and select the folder opened by Fame Codex. Select the folder that contains
manifest.json. - Click Open TH AI Pass, sign in to
de.aipass.netin Chrome, and open a chat page. Use the same Chrome profile where you installed the extension, and keep a TH AI Pass tab open while using Codex. - Return to Fame Codex and click Refresh models. Confirm that the status reads Extension connected and the model list is populated.
- Restart Codex if needed, choose a listed TH AI Pass model (its ID begins with
aipass/), and start a new Codex session.
If the extension is already installed, enable the integration, open your signed-in TH AI Pass chat tab, and refresh models. The launcher starts the bundled bridge automatically; no separate bridge server installation is required.
- Keep Fame Codex, the browser extension, and the signed-in TH AI Pass chat tab running during requests.
- If Chrome suspends the background chat tab, bring it to the foreground or exclude
de.aipass.netfrom Chrome's memory-saver settings. - Model availability and usage allowances depend on your TH AI Pass account. The launcher lists eligible chat/text models returned by the connected extension.
- Effort choices are loaded per model from TH AI Pass. Models with no advertised thinking levels use the web default;
maxappears as Extra High in Codex. After refreshing, restart Codex and open a new session to see updated choices. - If one computer lists fewer models, compare the models visible in the signed-in TH AI Pass website, use the same browser profile as the extension, and reload the extension before refreshing. Media and unsupported categories are intentionally excluded.
- The local prompt budget is 262,144 tokens, including serialized conversation history and tool definitions. This is a transport limit; individual upstream models may accept less context.
- Long sessions or large tool catalogs may exceed the prompt budget. Shorten the session or reduce enabled tools if this happens.
- This integration is experimental and uses a separate browser-extension transport from the ChatGPT MCP harness described below.
- Remote compaction is not supported for TH AI Pass sessions. Start a new session when switching to or from a TH AI Pass model.
- Requests are processed through TH AI Pass using your signed-in account. Use it only for content you are authorized to send to that service.
| Status or problem | What to do |
|---|---|
| Waiting for extension | Check that the extension is enabled at chrome://extensions, then reload the TH AI Pass chat tab and click Refresh models. |
| Extension connected, but no models | Confirm you are signed in and the chat page has loaded, then refresh models again. |
| Models missing from Codex | Confirm the launcher has a populated model list, then restart Codex. |
port_unavailable |
Another process is using local port 17843. Close the conflicting process if appropriate, then disable and re-enable TH AI Pass. Fame Codex leaves other processes running. |
| Prompt budget exceeded | Shorten the conversation or reduce enabled tools, then retry. |
To stop using the integration, disable Settings → TH AI Pass and restart Codex to refresh its model list.
Terminal install, updates & repair
Finish active Codex Web turns before quitting the launcher to update. Closing the bridge during a turn interrupts its connection. These installers select the platform and architecture, verify the published checksums, and preserve your ChatGPT profile and launcher settings.
macOS / Linux
curl -fsSL https://github.com/leonelfame/fame-codex/releases/latest/download/install-launcher.sh | shWindows PowerShell
irm https://github.com/leonelfame/fame-codex/releases/latest/download/install-launcher.ps1 | iexModels, modes & MCP setup
Automatic modes offer Luna/Think when the account has no reasoning selector; otherwise Instant–High, with Extra High and Pro available independently when exposed by the account.
| Mode | Sending messages | Local Codex tools |
|---|---|---|
| Browser-only | Automatic | No |
| Full harness (With Automation) | Automatic | Yes, through MCP |
| Zero Risk | Paste and send manually | Yes, through a separate MCP connector |
Zero Risk does not read or operate the ChatGPT page. Choose the model and Fame Codex Zero Risk connector yourself, paste and send the prepared prompt, then confirm Sent in the launcher. Automatic models ending in (Web) expose their supported Effort choices in Codex. Instant and each Pro version have separate entries to preserve their context budgets; older saved model entries keep their original fixed mode.
Full mode connects ChatGPT's tool calls back to the current Codex task through the official OpenAI tunnel-client. The tunnel is outbound: it does not expose a public IP, open an inbound port, or require router forwarding.
The launcher uses your system proxy for the tunnel. Explicitly configured proxy environment variables take priority; local connections stay local.
The launcher's MCP page guides the complete setup. For the exact clicks, see the video walkthroughs.
Limits
See Limits for the current ChatGPT message allowances for GPT-5.6 Sol Pro and GPT-6 Astra. Context limits depend on the account type and selected effort. Plus Medium/High uses a measured 90,000-token window. GPT-5.6 supports up to 270,000 tokens with experimental 3× context enabled, with native Codex compaction supported throughout.
GPT-6 Sol supports Bigger Context at Medium, High and available Extra High: 120,000 tokens on Plus, with compaction at 110,000, and 240,000 on Pro, with compaction at 220,000. GPT-6 Instant uses standard context. GPT-5.6 and GPT-6 Pro keep their existing Bigger Context limits.
- Finish the required setup, open MCP, create the Tunnel and regular API key, then press Connect harness.
- Enable ChatGPT Developer Mode and create a new Tunnel connector named exactly Fame Codex Native2, with Authentication: None and Allow all actions.
- Run Verify runtime to confirm that Fame Codex Native2 is attached and available.
Write/modify actions also require the ChatGPT workspace and its administrator policy to permit
them. See
developer mode and MCP apps.
Unexpected approval prompts fail closed unless --auto-approve-tool-calls is explicitly enabled;
that option clicks Allow once, never a permanent grant.
Diagnostics & subagents
Use Activity for safe local diagnostics and Settings → Run doctor for end-to-end health.
Use Settings → Updates → Check for updates to compare your installed version with published GitHub releases. Enable Include Beta releases to check prereleases too. Download update opens the installer in your browser; it does not close Fame Codex. Finish active Codex tasks before installing.
Settings can also cancel a retained browser turn or remove the Codex integration before uninstall.
Temporary Chat keeps new task conversations out of ChatGPT history and defaults on for new setups.
It is independent of New browser chat for each turn. Existing Save chats preferences migrate
to the equivalent Temporary Chat choice. Turn Temporary Chat off to keep ordinary saved conversations.
Settings → Trusted Codex homes lets ChatGPT chats recover their working folder from other Codex homes, such as an Orca account. Choose each folder deliberately: it must contain a sessions folder, at most seven can be added, and the change applies the next time the launcher-owned runtime starts.
Settings → Sleep Fame Codex temporarily restores your previous Codex route; Wake reconnects it.
Restart Codex after changing Sleep or refreshing the model catalog.
Set CODEX_CHATGPT_WEB_BROWSER_DIAGNOSTICS=1 only when every browser checkpoint needs a screenshot.
New installs use Compatibility V1 for cross-backend subagents. Native preserves Codex's own feature settings and enables plaintext Web-to-Web V2 delegation. Restart Codex and start a new task after changing the protocol:
codex-chatgpt-web subagents status
codex-chatgpt-web subagents compatibility-v1
codex-chatgpt-web subagents nativeRequirements & security
- This is unofficial browser automation, not an OpenAI API. ChatGPT UI changes can break selectors; drift fails explicitly instead of silently switching model or transport.
- Browser state is a sensitive login artifact, and the loopback listener is reachable by processes running as the same local user. Never share the launcher profile; use a trusted workstation.
- Release packages currently target macOS 13+ (Apple silicon/arm64 only; Intel Macs are not supported), Windows x64, and Linux x64. Runtime, tests, and packaging are gated on all three in CI; account-bound browser and MCP flows use the separate release validation.
- Builds are not yet platform-signed, so Gatekeeper or SmartScreen may warn. The installers verify the published SHA-256 manifest before installation.
Read the complete architecture and security model before enabling full mode. Report vulnerabilities through SECURITY.md.
Temporary Chat is a ChatGPT privacy mode; prompts are still processed by OpenAI.
Validation coverage: release validation.
This is independent software and is not affiliated with or endorsed by OpenAI. Use it only with your own account and in accordance with applicable Terms of Use and workspace policies; it does not bypass authentication or access controls.
Run from source & develop
git clone https://github.com/leonelfame/fame-codex.git && \
cd fame_codex && \
bun run appThis source path requires Bun 1.4.0. The command installs locked dependencies and opens the app.
bun run app
bun run dev:launcher
bun run src/cli.ts dev status
bun run dev:chat compaction-lab "Reply with exactly: DEV READY"
bun run verify
bun run smoke:subagents
bun run app:packagedev:launcher uses a separate profile and account under ~/.fame-codex-dev. dev:chat exercises the real browser and compaction paths with explicit simulated tool results, without changing your normal Codex route. See the DEV chat harness for setup and commands.
Leonelfame maintains and distributes Fame Codex at leonelfame/fame-codex.
See the Fame Codex project notice and license notices.