Execution Modes
| Mode | Description |
|---|
| Raita Managed | Generation runs on Raita's cloud pipeline. Billed in Raita tokens. No external API key required. |
| BYOK | Generation runs locally, calling AI provider APIs directly using your keys. API costs are charged to your provider account. |
Supported Providers
| Provider | Key required | Web search support | Notes |
|---|
| Raita Managed | Raita token wallet | Via managed pipeline | No external key needed |
| OpenAI | OpenAI API key | Yes — web search tool (/v1/responses) | GPT-4o, GPT-4.1, GPT-4.1 Mini, GPT-4o Mini, GPT-4.1 Nano, GPT-5.4 |
| Google Gemini | Gemini API key | Yes — via Gemini provider path | Gemini 2.5 Flash, Gemini 2.5 Pro, Gemini 3.1 Pro Preview, Gemini 3.1 Flash Lite Preview, Gemini 3 Flash Preview |
| Azure OpenAI | Azure endpoint + key + deployment name | Depends on deployed model | Any model deployed in your Azure resource |
| OpenRouter | OpenRouter API key | No | Access to many providers via single key (e.g., Kimi K2.5, Mercury 2) |
| Custom endpoint | Optional API key | No | Any OpenAI-compatible API (Ollama, LM Studio, etc.) |
Provider Configuration
All provider keys are configured in Settings → AI Provider.
| Provider | Fields required |
|---|
| Raita Managed | (none — uses token wallet) |
| OpenAI | API key |
| Google Gemini | API key |
| Azure OpenAI | Endpoint URL, API key, deployment name |
| OpenRouter | API key |
| Custom endpoint | Base URL, API key (optional) |
Model Pricing (Raita Managed)
Raita charges in Raita tokens (based on actual model API costs with ~3x margin). Last verified March 2026. Rates in tokens per 1M tokens (input / output):
OpenAI
- GPT-4.1 Nano: 1.25 / 5.0
- GPT-4o Mini: 3.75 / 15.0
- GPT-4.1 Mini: 10.0 / 40.0
- GPT-4.1: 50.0 / 200.0
- GPT-4o: 62.5 / 250.0
- GPT-5.4: 62.5 / 375.0
Google Gemini
- Gemini 3.1 Flash Lite Preview: 2.5 / 10.0
- Gemini 2.5 Flash: 7.5 / 62.5
- Gemini 3 Flash Preview: 12.5 / 75.0
- Gemini 2.5 Pro: 31.25 / 250.0
- Gemini 3.1 Pro Preview: 50.0 / 300.0
OpenRouter (sample)
- Kimi K2.5: 11.25 / 55.0
- Mercury 2: 6.25 / 18.75
Notes
{webSearch} behavior varies by provider. See Scraper Macros Reference for the per-provider breakdown.- For BYOK providers, model selection can be overridden per-prompt in the prompt editor's Model field.
- Azure OpenAI available models depend on which models you have deployed in your Azure resource.
- For Raita Managed mode, the model catalog is updated regularly as new models and pricing become available.
