Agents
Setup guide

Cline, Kilo Code and Roo Code with an uncensored model

Cline, Kilo Code and Roo Code all take an uncensored model through their OpenAI Compatible provider, with the same settings in each.

Updated
FormatOpenAI Chat Completions
Endpoint/chat/completions
Configured inExtension settings
Switch withthe extension settings
Context1,048,576

Which model to use in Cline, Kilo Code, Roo Code

Each profile takes one model id, so keep DeepSeek V4.1 Flash 1017 for everyday work and a second profile on Kimi K3.1 Pro for hard tasks.

ModelUse it forModel id
Kimi K3.1 ProHard multi-file and security workkimi-k3.1-pro-uncensored
DeepSeek V4.1 FlashEveryday work, and the default heredeepseek-v4.1-flash-1017-uncensored
GLM-5.3A second opinion, text onlyglm-5.3-uncensored
GLM-5.3 FlashBulk edits and many subagentsglm-5.3-flash-uncensored

Set it up

SettingValue
API ProviderOpenAI Compatible
Base URLhttps://api.silkcompute.com, or https://api.silkcompute.com/v1 if the extension asks for /v1
API keyYour key from Silk Compute
ModelA model id, for example deepseek-v4.1-flash-1017-uncensored
Context window1048576, if asked
  1. Open the extension's settings

    In VS Code, open Cline, Kilo Code or Roo Code and go to its API provider settings.

  2. Choose OpenAI Compatible

    Set API Provider to OpenAI Compatible, which Cline documents as the provider for "APIs compatible with the OpenAI API standard" (Cline docs).

  3. Fill in the settings

    Use the values in the table above, with any of these model ids.

    Model ids
    kimi-k3.1-pro-uncensored
    deepseek-v4.1-flash-1017-uncensored
    deepseek-v4.1-flash-1017-fast-uncensored
    glm-5.3-uncensored
    glm-5.3-flash-uncensored
    hy4-uncensored
  4. Start a task

    If the extension reports a context error, set the context window to 1048576.

Other APIs that work with Cline, Kilo Code, Roo Code

Any Chat Completions API connects, and three of the seven APIs for coding agents have a 1M-token context.

With a local model

For a local server, enter its address as the base URL and one of the local models as the model id.

Silk Compute's own guide is at silkcompute.com/docs/other-tools.