Skip to main content
Configuration

API Keys & Providers

Connect CogTog to 15+ AI providers. Choose from cloud services like Anthropic, OpenAI, Google, and OpenRouter, or run models locally with Ollama.

LLM Providers

Cloud Providers

Connect to cloud AI providers for access to the most powerful models. Each provider requires an API key.

OpenRouter

Recommended

Access 200+ models from multiple providers with a single API key. Best value and flexibility.

Claude
GPT-4
Gemini
Llama
Mistral
200+ more

Getting your API key:

  1. 1. Visit openrouter.ai
  2. 2. Sign up for a free account
  3. 3. Go to Keys → Create new key
  4. 4. Copy the key and paste it in CogTog Settings
Pay-as-you-go pricing • No subscription required

Anthropic

Direct access to Claude models (Opus 4, Sonnet 4, Haiku). Best for advanced reasoning and coding tasks.

Claude Opus 4
Claude Sonnet 4
Claude 3.5 Sonnet
Claude Haiku

Getting your API key:

  1. 1. Visit console.anthropic.com
  2. 2. Sign up and verify your account
  3. 3. Add billing information
  4. 4. Create an API key in the dashboard

OpenAI

Access GPT-4o, o1, and o3-mini models. Industry-leading models for general-purpose tasks.

GPT-4o
GPT-4o Mini
o1
o3-mini

Getting your API key:

  1. 1. Visit platform.openai.com
  2. 2. Create an account or sign in
  3. 3. Set up billing
  4. 4. Go to API Keys and create a new key

Google AI

Gemini models with massive context windows (up to 2M tokens). Great for processing large documents.

Gemini 2.0 Flash
Gemini 2.0 Thinking
Gemini 1.5 Pro

Getting your API key:

  1. 1. Visit aistudio.google.com
  2. 2. Sign in with your Google account
  3. 3. Click "Get API Key"
  4. 4. Copy and use in CogTog
Generous free tier available

15+ Providers Supported

CogTog supports all major AI providers, giving you flexibility and choice:

Anthropic (Claude)
OpenAI (GPT-4o)
Google (Gemini)
Groq
Mistral
DeepSeek
Cohere
Together AI
Fireworks AI
OpenRouter
Ollama (local)
Custom endpoints

Local Models with Ollama

Run AI models locally on your computer with Ollama. No API keys, no internet required, completely free.

Ollama

Local
Free

Run over 3,400 open-source models locally. Perfect for privacy-sensitive work or when working offline.

Popular Models:

Llama 3.3 70B
Meta's latest, very capable
DeepSeek Coder V2
Excellent code generation
Mixtral 8x7B
Fast MoE model
Qwen 2.5 Coder
Strong coding from Alibaba

Setting up Ollama:

  1. 1. Download Ollama from ollama.ai
  2. 2. Install and run Ollama
  3. 3. Open terminal and run: ollama pull llama3.3:70b
  4. 4. CogTog will automatically detect Ollama and show available models
100% private - your data never leaves your machine
No API costs - completely free
Works offline - no internet needed
3,400+ models available from the community

Adding API Keys to CogTog

1

Open Settings

Click the gear icon in the top-right corner or press Ctrl+,

2

Go to Providers

Navigate to the "Providers & Models" section

3

Select Provider

Choose the provider you want to configure (OpenRouter, Anthropic, OpenAI, etc.)

4

Paste API Key

Paste your API key and click "Save"

5

Test Connection

CogTog will test the connection and show available models

API Key Security

Your API keys are stored securely and encrypted on your local machine. CogTog never sends your keys to our servers.

Keys are encrypted at rest
Never transmitted to CogTog servers
Only used to communicate with AI providers

Which Provider Should You Choose?

For Best Value & Flexibility

Choose OpenRouter. Access 200+ models with one API key, pay-as-you-go pricing.

Recommended for most users

For Best Performance

Choose Anthropic for Claude models. Best for complex reasoning, coding, and long conversations.

For Privacy & Offline Work

Choose Ollama. Run models locally, no API costs, complete privacy.

For Large Documents

Choose Google AI for Gemini. Massive 2M token context window for processing large files.

Learn More