AI Chat Models
MachinaOs supports three major AI providers for chat completions.Available Models
Models are fetched dynamically from each provider’s API when you add your API key.OpenAI
| Model | Best For |
|---|---|
| gpt-5.2 | Most capable, professional tasks |
| gpt-5 | Advanced reasoning, multimodal |
| o3 | Complex reasoning |
| o4-mini | Fast, cost-efficient reasoning |
Anthropic Claude
| Model | Best For |
|---|---|
| claude-opus-4.5 | Best for coding and agents |
| claude-sonnet-4.5 | Balanced performance |
| claude-haiku-4.5 | Fast responses |
Google Gemini
| Model | Best For |
|---|---|
| gemini-3-pro | Most intelligent, complex tasks |
| gemini-3-flash | Fast, frontier performance |
| gemini-2.5-pro | Long context, multimodal |
Adding API Keys
- Click the key icon in the toolbar
- Select the provider
- Enter your API key
- Click Validate to test
API keys are encrypted and stored locally. They’re never sent to MachinaOs servers.
OpenAI Chat Model
Parameters
The model to use (gpt-5.2, gpt-5, o3, o4-mini)
The message to send. Supports template variables.
Randomness (0 = deterministic, 1 = creative)
Maximum response length
Output format: text or json_object
Output
Example
Anthropic Claude Model
Parameters
Claude model (claude-opus-4.5, claude-sonnet-4.5, claude-haiku-4.5)
The message to send
System instructions for the model
Randomness (0-1)
Maximum response length
Output
Example
Google Gemini Model
Parameters
Gemini model (gemini-3-pro, gemini-3-flash, gemini-2.5-pro)
The message to send
Randomness (0-1)
Maximum response length
Content safety level
Output
Comparing Providers
| Feature | OpenAI | Claude | Gemini |
|---|---|---|---|
| Speed | Fast | Medium | Fast |
| Reasoning | Good | Excellent | Good |
| Context Window | 128K | 200K | 1M |
| Multimodal | Yes | Yes | Yes |
| JSON Mode | Yes | No | No |
Common Use Cases
Text Generation
Data Extraction
Translation
Summarization
Tips
Error Handling
| Error | Cause | Solution |
|---|---|---|
| 401 Unauthorized | Invalid API key | Check/update API key |
| 429 Rate Limited | Too many requests | Add delay, reduce frequency |
| 500 Server Error | Provider issue | Retry later |