Transition to DeepSeek-V4 Model Names for Long-Term Stability
Update your model IDs to avoid service disruption by July 2026.
Switch your API calls from 'deepseek-chat' to 'deepseek-v4-pro' or 'v4-flash' to access the 1.6T parameter model and avoid the July 2026 deprecation.
The Scenario
You are a developer using DeepSeek for coding assistance in VS Code and want to ensure your setup is using the latest flagship models before the legacy API names are retired.
Before & after
Users would manually check for new endpoints or wait for third-party integrations to update their dropdown menus, often taking 15-20 minutes to find documentation.
Update your model parameter to `deepseek-v4-pro` or `deepseek-v4-flash` while keeping your existing API base URL. This switch takes approximately 2 minutes.
The Prompt
In your application configuration or VS Code 'Continue'/'Cline' settings, update the model string: "model": "deepseek-v4-pro" or "model": "deepseek-v4-flash"
The legacy names 'deepseek-chat' and 'deepseek-reasoner' are deprecated and will be discontinued on July 24, 2026. Transitioning now ensures your production apps or IDE extensions don't break when the legacy names are retired.
Source
Change Log | DeepSeek API Docs"The model parameter should be set to deepseek-v4-pro or deepseek-v4-flash. The two legacy API model names... will be discontinued."