Pricing Calculator
Estimate your Gemini 3.5 Flash costs by thinking_level. Includes the cost difference vs the old gemini-3-flash-preview default (thinking_level=high).
Configure your usage
~4,000 thinking tokens per call
10050K
1008K
110K
Your config (medium)
$48.30
/month
$0.016100 per call
~4,000 thinking tokens/call
vs old default (high)
−$42.00
/month
Old default (gemini-3-flash-preview): $90.30/mo
vs new default (medium)
−$0.00
/month
New default (gemini-3.5-flash): $48.30/mo
Cost breakdown (per call)
Input tokens (1,000 tokens)$0.000350
Output tokens (500 tokens)$0.001750
Thinking tokens (medium) (4,000 tokens)$0.014000
Total per call$0.016100
Estimates only. Thinking token counts vary by prompt complexity. Verify pricing at ai.google.dev. Not affiliated with Google.
Gemini 3.5 Flash pricing reference
| Token type | Price per 1M tokens | Notes |
|---|---|---|
| Input tokens | $0.35 | User prompt + system prompt + context |
| Output tokens | $3.50 | Model response text |
| Thinking tokens | $3.50 | Internal reasoning tokens (not in output) |
| Cached input tokens | $0.0875 | 75% discount via context caching |
Prices as of June 2026. Verify at ai.google.dev for current rates. Not affiliated with Google.
Thinking token estimates by level
minimal
~1,000 thinking tokens
low
~2,000 thinking tokens
medium
~4,000 thinking tokens
high
~8,000 thinking tokens
These are estimates based on typical observed usage. Actual thinking token count varies by prompt complexity.