Opencode + Antigravity oAuth Fixes EVERYTHING! A New Alternative to Claude Code Max FOR FREE!

Key Concepts

• Oath Spoofing: A technique used to bypass subscription restrictions by mimicking legitimate authentication processes.
• Open Code: An open-source CLI (Command Line Interface) for interacting with various LLMs (Large Language Models).
• Claude (Opus, Sonnet, 4): LLMs developed by Anthropic, offering different levels of performance and cost.
• Anti-Gravity: A service offering free (rate-limited) access to LLMs, used as a workaround.
• Gemini: LLMs developed by Google, integrated with Anti-Gravity and accessible through Open Code.
• Rate Limiting: Restrictions on the number of requests a user can make within a given timeframe.
• Multi-Account Rotation: A technique to bypass rate limits by automatically switching between multiple accounts.
• CLI (Command Line Interface): A text-based interface for interacting with computer programs.

Anthropic’s Safeguards & Workarounds for LLM Access via Open Code

Introduction & Anthropic’s Actions

On January 9th, Anthropic implemented stricter technical safeguards during a tools update, effectively blocking third-party tools like Open Code from utilizing Claude Pro and Max subscription credentials through oath spoofing. This action targeted developers who were leveraging Open Code to access models like Claude 4, Opus, and Sonnet at reduced, flat rates – typically around $200 per month. Anthropic viewed this practice as unauthorized client spoofing, an abuse of consumer plans intended for official Cloud Code usage, and a threat to their business model. The immediate consequence was broken logins, errors, and temporary account restrictions for users.

Anthropic’s Rationale & Future Plans

The speaker acknowledges Anthropic’s position, stating, “I totally understand where Anthropic is coming from cuz it's clearly bleeding money without safeguards in place.” Anthropic aims to encourage usage within its own ecosystem, particularly in anticipation of launching its own IDE (Integrated Development Environment) and more advanced developer tooling. The speaker believes this move is “justifiable for them to stop the bleeding of outflow of money and to keep odd products within their own ecosystem.”

OpenAI’s Response & Ecosystem Support

Interestingly, OpenAI reacted to Anthropic’s move by publicly affirming their commitment to supporting a thriving ecosystem of generic coding tools. OpenAI employees reportedly expressed their “100% invested” stance, allowing developers to use Codeex (OpenAI’s coding model) within Open Code using a Codeex plan, rather than implementing similar restrictions.

Utilizing Anti-Gravity & Gemini as a Workaround

The Anti-Gravity Solution

Despite the limitations imposed on Claude access, a free workaround exists using Anti-Gravity’s free tier, integrated within Open Code. While Anti-Gravity is rate-limited, the speaker emphasizes that “everything is rate limited,” and this is a common constraint. The workaround involves utilizing Anti-Gravity’s subscription oath through Open Code, granting access to Claude and other state-of-the-art models. Once the rate limit is reached, the system automatically switches to the Gemini model.

Step-by-Step Implementation

1. Terminal Access: Open your preferred terminal or coding agent (e.g., Cursor, Cloud Code).
2. Plugin Installation: Copy and paste the command (provided in the video description and repo link) into the terminal to install the Anti-Gravity model definition plugin for your agent.
3. Manual Configuration (Alternative): Alternatively, manually add the plugin to your Open Code configuration file (JSON format), requiring authentication with O login and model addition. Specific configuration details are provided in the video.
4. Model Access: After installation, open Open Code and use the /models command to search for and select Anti-Gravity.
5. Multi-Account Rotation: Configure multiple Google accounts for automatic rotation to bypass rate limits.

Gemini Integration

Since Gemini subscriptions are also integrated with Open Code, users can seamlessly switch to and utilize Gemini when Anti-Gravity’s rate limit is reached.

Additional Model Options & Alternatives

OpenAI Integration

Users can directly integrate their OpenAI plan (ChatGPT Pro or Plus) within Open Code by switching to the OpenAI provider tab and providing their credentials.

Miniax & ZAI’s GLM 4.7

• Miniax: Offers a cheaper alternative to GPT and Opus, with impressive generation quality achieved through training on outputs from models like Gemini 3.0 and Opus. Details on setup are available in the video description.
• ZAI’s GLM 4.7: Another option, with the fifth iteration of the GLM model series expected soon. Their coding plan can also be used directly within Open Code.

The speaker advises users to “choose the one that's best fitted for you,” acknowledging that the optimal solution depends on individual needs and preferences.

Community Resources & Future Outlook

Resources:

• World of AI Newsletter: A weekly newsletter providing updates on the AI space (link in description).
• Video Description: Contains links to the Open Code repo, Anti-Gravity setup instructions, Miniax integration details, and other relevant resources.
• Private Discord: Offers access to multiple AI tool subscriptions, daily news, and exclusive content (paid membership).

Future Predictions

The speaker anticipates that the current workaround may be short-lived, predicting that Anthropic will likely find ways to block access through Google or other methods. However, they believe this situation will “force competition to step up and ship better and faster products,” ultimately benefiting users with improved plans and outputs.

Conclusion

Anthropic’s recent safeguards, while disruptive in the short term, are a strategic move to protect their business model and encourage ecosystem lock-in. While direct access to Claude through Open Code is currently restricted, viable workarounds exist using Anti-Gravity, Gemini, and alternative models like Miniax and ZAI’s GLM 4.7. The situation highlights the dynamic nature of the LLM landscape and the ongoing competition between providers to deliver value to developers and users. The speaker ultimately believes this will lead to innovation and better offerings in the long run.