SI-APIPortuguês (BR)Open console Model IDs and input support
Copy the model ID into your API request. Coding tool configurations already include the full list.
| Model / API ID | Guide input | Reasoning controls | Image admission |
|---|---|---|---|
GLM 5.3 Flashglm-5.3-flash | Text | low, high, max | Enabled in gateway configuration |
DeepSeek V4.1 Flashdeepseek-flash | Text | none, low, high, max | Enabled in gateway configuration |
Kimi K3.1 Prokimi-k3.1-pro | Text | low, high, max | Enabled in gateway configuration |
Kimi K3.1 Pro HighSpeedkimi-k3.1-pro-highspeed | Text | low, high, max | Enabled in gateway configuration |
GLM 5.3glm-5.3 | Text | low, high, max | Disabled in gateway configuration |
MiniMax M3.1 Flashminimax-m3.1 | Text | Model default | Not configured for this API ID |
Qwen3.8 Maxqwen3.8-max | Text | none, low, medium, xhigh | Enabled in gateway configuration |
Agent compatibility — Client versions, tested reasoning fields and known limits.
Model configuration
The setup guides cover text, streaming and client-defined function tools. MiniMax M3.1 Flash uses its default reasoning behavior; M3 settings do not apply.
Image input requires both an enabled model route and a client that sends images. The downloadable Agent configurations use text input.
SI-API accepts output limits up to 32,768 tokens per request. Model routes may have lower limits.