GLM-5.3-Flash Is Free on OpenCode Zen for a Limited Time With 200K Context

OpenCode announced on October 16 that GLM-5.3-Flash is now free. Zen's model list has a new GLM-5.3-Flash Free (model ID glm-5.3-flash-free) that you can call as soon as you sign in to Zen.

What you get

Input, output, and cached reads on GLM-5.3-Flash Free all cost nothing. It is free for a limited time, and no end date has been announced yet. The provider follows a zero-retention policy and does not use your requests to train models.

The free version has a 200K-token context and up to 32K output tokens per request. The pay-as-you-go GLM-5.3-Flash supports a 1M-token context and costs $0.15 per million input tokens and $0.50 per million output tokens. The free version covers everyday coding and agent work; switch to the paid one when you need to load a very large codebase in one go.

Model specs

GLM-5.3-Flash is an open-weights model from Zhipu AI released in late August, built for efficient coding and long-horizon agent tasks. It reads text, images, video, and PDFs, supports tool calling, structured output, and reasoning, and lets you pick low, high, or max reasoning effort.

How to use

1. Open OpenCode, run /connect to sign in to Zen and paste your API key, then use /models to switch to GLM-5.3-Flash Free. You can also set the model to opencode/glm-5.3-flash-free in your config.
2. For direct API calls, use the endpoint https://opencode.ai/inference/openai/v1/chat/completions with the model name glm-5.3-flash-free in the OpenAI-compatible format.

See the OpenCode Zen docs and OpenCode's announcement on X. DeepSeek V4.1 Flash Free is also free on Zen right now if you want to compare a few models.