Google Gemini
Google Gemini offers industry-leading context windows (up to 1M tokens) with strong multimodal capabilities. Great for analyzing large codebases, documents, and images.
Setup
# 1. Get your API key from https://aistudio.google.com/apikey
export GEMINI_API_KEY=...
# 2. Verify detection
arcana doctor
# 3. Run
arcana run "hello"
Available Models
| Model | Context | Best For |
|---|
gemini-2.5-pro | 1M | Complex reasoning, coding, analysis |
gemini-2.5-flash | 1M | Fast, cheap, good for most tasks |
gemini-2.0-flash | 1M | Previous gen, still very capable |
Config File
{
"provider": "google",
"model": "gemini-2.5-pro",
"utilityModel": "gemini-2.5-flash"
}
Environment Variables
| Variable | Description |
|---|
GEMINI_API_KEY | Your Google AI API key (required) |
GOOGLE_GENERATIVE_AI_API_KEY | Alternative env var name |
Pricing Notes
| Model | Input | Output |
|---|
| gemini-2.5-pro | $1.25 | $10.00 |
| gemini-2.5-flash | $0.15 | $0.60 |
Prices per 1M tokens. Check ai.google.dev/pricing for current rates.
Tips
- Gemini's 1M context window is perfect for analyzing entire codebases or long documents
gemini-2.5-flash is extremely cost-effective for everyday tasks- Gemini supports multimodal input — images, video, and audio
- Free tier available with rate limits
Last updated: Jul 24, 2026
ARCANA