Rate transparency — costs are calculated as:
token count from API response ×
hardcoded rate below.
Token counts are authoritative (from Google). Rates are from training data and
may have changed.
Always verify at
ai.google.dev/pricing and update
llos/cost_tracker.py → RATES dict if they differ.
gemini-2.0-flash: $0.075/1M in · $0.30/1M out
gemini-1.5-flash: $0.075/1M in · $0.30/1M out
gemini-1.5-pro: $3.50/1M in · $10.50/1M out