Accessing GLM-5.2 for Free via OpenCode

AICodeKing2026-06-24go watch the original →

the gist

The OpenCode platform provides free access to the GLM-5.2 coding model under the alias 'Big Pickle', offering approximately 200 requests per 5-hour window for non-sensitive development tasks.

Accessing the Model

Users can access the GLM-5.2 model for free by installing the OpenCode CLI and selecting the 'Big Pickle' model alias. The installation can be performed via terminal using curl -fsSL https://opencode.ai/install | bash or npm install -g opencode-ai. After authenticating with OpenCode Zen, users must run the /models command and explicitly select 'Big Pickle' to avoid using paid endpoints.

Operational Workflow

To maximize the 200-request limit, users should adopt a structured approach to agentic coding tasks.

Start in 'Plan' mode to have the model inspect the repository and outline necessary changes before switching to 'Build' mode for implementation.
Use scoped prompts that explicitly instruct the model to inspect files, explain current implementations, run tests, and perform type checks.
Avoid broad, open-ended requests that may trigger excessive background model calls and exhaust the 5-hour request quota prematurely.

Usage Constraints

Because the free endpoint is provided as a stealth service, the underlying model and availability are subject to change without notice. Furthermore, OpenCode reserves the right to use data processed through the free 'Big Pickle' endpoint for model improvement. Consequently, this service should be restricted to open-source projects, personal experiments, and non-sensitive code. It is strictly advised against using this endpoint for private company code, credentials, or customer data.