Choosing an AI Provider
Batch Generate Text with any AI works with OpenAI-compatible chat APIs. Nine provider presets are built in, and you can also use a custom OpenAI-compatible endpoint.
Quick recommendation
| If you want… | Use |
|---|---|
| Cheapest option for bulk generation | DeepSeek |
| Best overall reliability and quality | OpenAI |
| Fast responses | Cerebras or Fireworks AI |
| Access to many models from one key | OpenRouter |
| European provider | Mistral AI |
| Generous free quota to start | Google AI Studio |
| Open-source model selection | Together AI |
All supported presets
| Provider | Default model | Strengths | Sign up |
|---|---|---|---|
| OpenAI | gpt-4o-mini | Best reliability, broad model range, strong general quality | platform.openai.com |
| DeepSeek | deepseek-chat | Extremely low cost and strong value for bulk generation | platform.deepseek.com |
| OpenRouter | deepseek/deepseek-chat | Routes to many models from one API key | openrouter.ai |
| Mistral AI | mistral-large-latest | Strong multilingual support, EU-based provider | console.mistral.ai |
| Together AI | meta-llama/Llama-3.3-70B-Instruct-Turbo | Wide open-source model selection, competitive pricing | api.together.ai |
| xAI (Grok) | grok-3-mini | Grok models by xAI | console.x.ai |
| Fireworks AI | accounts/fireworks/models/llama-v3p3-70b-instruct | Fast, cost-effective open-source inference | fireworks.ai |
| Cerebras | llama-3.3-70b | Very fast inference on Cerebras hardware | cloud.cerebras.ai |
| Google AI Studio | gemini-2.0-flash | Fast Gemini models, large context windows, generous free tier | aistudio.google.com |
Cost considerations
All providers charge per token - roughly 1 token per 0.75 words. For content generation:
- An 800-word blog article often uses roughly 1,500-2,500 tokens total, including prompt and output
- A batch of 20 articles often uses roughly 30,000-50,000 tokens
- Low-cost providers can make bulk generation very inexpensive
- Free tiers may have strict rate limits even when token pricing is low
The app displays token usage per item and run totals so you can track costs.
Using a custom endpoint
If you use a self-hosted model, LM Studio, LocalAI, Azure OpenAI, or a provider not in the preset list, select Custom (OpenAI-Compatible) from the provider dropdown and enter your endpoint’s Base URL manually.
The endpoint must support OpenAI-compatible /chat/completions requests. See Finding Your Base URL for details.
Switching providers
You can switch providers at any time from the key icon in the header.
To keep more than one provider ready to go, save them as separate connections. In AI Provider Settings, use the connection dropdown and choose Add new connection. Each connection has its own name, provider, Base URL, model, and encrypted API key.