AI Models

Choose from multiple AI providers and models, or bring your own API key for maximum flexibility.

Choose Your AI Model

Thoughtful gives you the freedom to work with the AI model that best fits your needs. Whether you prefer Claude's thoughtful reasoning, GPT's versatility, or Gemini's speed, you can switch between providers without changing how you work.

Supported Models

Thoughtful supports leading AI models from three major providers:

Anthropic Claude

  • Claude Opus 4.6 - Our default model, excellent for complex reasoning and nuanced tasks
  • Claude Sonnet 4.5 - A great balance of capability and speed
  • Claude Haiku 4.5 - Fast responses for simpler tasks

OpenAI GPT

  • GPT-5.4 - Latest GPT-5 generation with stronger coding and long-context performance
  • GPT-5.2 - Powerful general-purpose model with a large context window
  • GPT-5.2 Pro - Enhanced capabilities for demanding tasks

Google Gemini

  • Gemini 3 Flash - Extremely fast responses with a massive context window
  • Gemini 3 Pro - Google's flagship model for complex tasks

Switching Between Models

Changing your AI model is simple:

  • Look for the model picker below the chat input
  • Click to open the dropdown menu
  • Select your preferred model
  • Your choice is remembered for future conversations

Each model has different strengths. You might use Claude Opus for detailed analysis, switch to Gemini Flash for quick questions, or try GPT-5.4 for a fresh perspective on a problem.

Using Your Own API Key

By default, Thoughtful provides AI access through our platform. However, you can connect your own Anthropic API key to use your own credits:

  • Go to Settings > Integrations
  • Find the Anthropic API Key card under Personal Integrations
  • Click Connect and enter your API key
  • Your key is encrypted and stored securely

Benefits of using your own API key:

  • Cost control - Pay Anthropic directly and track your usage
  • Higher limits - Avoid any platform rate limits
  • Enterprise features - Access any enterprise features tied to your Anthropic account

To disconnect your API key, return to Settings and click Disconnect. You will automatically fall back to the platform's shared AI access.

Skills Work Across All Models

One of Thoughtful's most powerful features is that skills are model-agnostic. When you install a skill or use built-in capabilities, they work identically regardless of which AI model you have selected.

This means:

  • Consistent behavior - The same skill produces reliable results whether you are using Claude, GPT, or Gemini
  • Easy experimentation - Try different models to see which works best for your workflow
  • Future-proof - As new models become available, your existing skills continue to work

Skills define what the AI can do (tools, integrations, workflows), while the model determines how the AI reasons and responds. This separation keeps your workspace portable and flexible.

Why Model Flexibility Matters

Being locked into a single AI provider creates risk:

  • Provider outages - If your only provider has issues, your work stops
  • Pricing changes - You have no leverage if costs increase
  • Capability gaps - Different models excel at different tasks
  • Innovation pace - New providers and models emerge regularly

Thoughtful's multi-model approach means you are never stuck. If a new model launches that better fits your needs, you can switch in seconds. Your pages, skills, integrations, and workflows all continue working exactly as before.

This is not just about having options. It is about ensuring your AI-powered workspace remains productive, cost-effective, and at the cutting edge, no matter how the AI landscape evolves.