Gemini 4 Carbon Pricing

Deep multi-step reasoning for hard coding, math, and science.

Estimate your cost

Pick what one request looks like. You pay only for what you run: no minimums, no commitments.

Estimated cost

50% off until 30 Nov 2026

Was $0.03Now $0.015

per request

1,000 input tokens at $2.5 per million
$0.0025
1,000 output tokens at $12.5 per million
$0.0125

Rates

Pay as you go

50% off until 30 Nov 2026

Partner model served through the Segmind API and billed in Segmind credits

Input
Was $5.000 now $2.500
Output
Was $25.000 now $12.500
per million tokens
Reasoning effort
Thinking is billed as output tokens at the rate above. The effort parameter changes how many of them the model spends, not the rate, so the same prompt costs more at a higher level.
  • No upfront costs - Only pay for what you use
  • Partner-hosted - Served from the provider's own API through Segmind
  • Universal credits - Use anywhere on Segmind
  • No setup - Call the API right away