Chat Interface
Chat with your fine-tuned models directly in the browser.
The chat interface is the fastest way to interact with your fine-tuned models. Once training completes, click Open in chat on any model card.
Layout
The chat page has three areas:
Model info (top)
Shows the model's display name, base model, and your use-case description. This helps you remember what each model was trained for when switching between them.
Message area (center)
Your conversation with the model. Messages appear in a familiar chat layout — your messages on the right, the model's responses on the left.
Composer (bottom)
A text input where you type messages. Press Enter to send. The model streams its response in real-time.
Chat history (left sidebar)
A list of all saved conversations with the current model. Click any conversation to resume it.
How it works
When you send a message:
- Your message is sent to Commissioned's backend
- The backend routes it to the fine-tuned model (on OpenAI, Google, or Commissioned's GPU cluster)
- The model's response streams back in real-time
- The full conversation is saved automatically
The chat uses the same infrastructure as the API. There's no difference in model quality between chat and API access — they hit the same fine-tuned model.
Rate limits
Chat messages are rate-limited based on your plan:
| Plan | Limit |
|---|---|
| Free | Standard (per 3-hour window) |
| Pro | 10x standard |
| Enterprise | Custom |
When you hit the limit, a prompt appears to upgrade or wait for the window to reset. See Rate Limits for details.