As recorded at collection: opencode.ai Zen, setup for using free models
As recorded at collection: This is a setup method for using the deepseek-v4-flash-free model for free on the AI model service opencode.ai Zen, without separate registration or payment. To connect, you must enable local routing and enter the request URL, the public authentication value public, the model name, and the request override value.
- What you get
- Free deepseek-v4-flash-free access
- Price
- No payment
- Valid until
- To be confirmed
- Requirements
- Local routing must be enabled · enter public in the API key field · {"x-opencode-client": "desktop"} must be entered in the request override field · no usage period or usage limit information provided
Details
This is a post sharing the setup for connecting to Zen's free model operated by opencode.ai. It uses a method of entering the public authentication value without a separate application or payment process, and includes two screenshots of the settings screen.
In the settings of the program you use, enable local routing and add a new provider. Enter https://opencode.ai/zen/v1 in the request URL field and public in the API key field, which is the authentication string required to use the service. Specify deepseek-v4-flash-free in the actual request model field. You must enter {"x-opencode-client": "desktop"} in the local request override field.
If any one of the request URL, API key, model name, or request override value is missing, the connection will not work, so they must be entered exactly as shown. No separate usage period or usage limit has been announced. Because this method uses a public authentication value, access may be blocked if the service policy changes.
Sources and verification history
- Collection channel
- linux.do
- Collected at
- August 6, 2026, 10:20 KST
- First published
- Updated
- Source
- linux.do/t/topic/2711359