LLM Providers
Set up AI model providers like OpenAI, Anthropic, and Google.
LLM Providers
LLM providers are the AI companies whose models power your agents. FormWise does not lock you into a single provider - you can pick the best model for each task, from any of the supported providers.
What are LLM providers?
Every time your agent generates a response, it uses a large language model (LLM) behind the scenes. These models are built and hosted by companies like OpenAI, Anthropic, and Google. Every model in the Model dropdown works out of the box and runs on your plan's included credits. If you add your own API key for a provider, calls to that provider's models run on your key instead, and the provider bills you directly rather than FormWise deducting credits.
Plan availability
Saving your own provider keys is available on the Pro and Agency plans. Keys you saved before a plan change stay connected.
Supported providers
FormWise currently supports connecting your own API key for three model providers, each with the latest models from that company plus recent prior generations:
OpenAI
OpenAI offers the GPT family of models, including GPT-5.5 and GPT-4o, alongside smaller, faster options like GPT-4o Mini for everyday tasks on a budget.
Get your API key at platform.openai.com.
Anthropic Claude
Anthropic's Claude models are known for careful reasoning and long-context handling - currently the Claude Opus family for the most demanding work, Claude Sonnet as a balanced default, and Claude Haiku for fast, everyday tasks.
Get your API key at console.anthropic.com.
Google Gemini
Google's Gemini models offer strong reasoning and multimodal capabilities - including Gemini 3.1 Pro for advanced, multi-step reasoning and Gemini 3 Flash for fast, cost-efficient everyday use.
Get your API key at aistudio.google.com.
The same page also takes an ElevenLabs key, which powers voice-to-voice speech for your agents rather than text generation.
Model lineups move quickly. The Model dropdown inside each Agent node always shows the current, live set of models - treat the names above as examples, not an exhaustive list.
Adding an API key
- Go to Settings and open Integrations.
- Stay on the API Keys tab.
- Find the card for the provider you want to connect.
- Paste your API key into the API Key field.
- Click Connect.
That's it - the card shows a Connected badge, and calls to that provider's models across all your Flows now run on your key. To remove a key, click Disconnect on the card; to replace one, disconnect it and connect the new key.
Choosing a model for your agent
To select which model your agent uses:
- Open your Flow in the workflow builder.
- Click on an Agent node.
- In the node settings, find the Model dropdown.
- Choose your preferred provider and model.
How to decide which model to use
- Need top-quality results? Go with a flagship model like GPT-5.5, Claude Opus, or Gemini 3.1 Pro.
- Want fast, affordable responses? Try a lighter model like GPT-4o Mini, Claude Haiku, or Gemini 3 Flash.
- Working with images or files? Use a multimodal model - most current flagship and mid-tier models across all three providers support this.
- Not sure? Start with a balanced mid-range model like Claude Sonnet or GPT-4o and adjust based on your results.
Different agents within the same Flow can use different models, so you can mix and match to balance cost and quality.