Gemini 3.7 Flash pricing
$0.75 / $3.75, then it doubles
Official Gemini Developer API: Standard intro $0.75 in / $3.75 out per million tokens (output includes thinking) through 31 Dec 2026. From 1 Jan 2027: $1.50 / $7.50. QCode does not currently offer the Gemini series; please use Claude, GPT or Chinese models.
Updated 2026-10-08
Prices change: on QCode, Claude, GPT and Chinese models are billed per token — live rates on /models.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard / AMEX), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
The rates that matter
Standard intro
Per million tokens. Output includes thinking tokens. Through 31 Dec 2026.
From 1 Jan 2027
Intro ends. Same list as 3.6 Flash today — and the two already share the intro rate.
Batch / Flex
Official Batch and Flex are half of Standard during intro. Priority intro is $1.35 / $6.75.
Context cache
Intro $0.075 per million, plus $0.50 / million / hour storage. Both double in 2027.
This page is only the bill
Specs live on /gemini-3-7-flash-guide. Here we only read the gemini-3.7-flash table on Google’s pricing page. Free tier is free of charge; the numbers below are Paid. Search / Maps grounding is separate: 5,000 free queries a month, then $14 / 1,000.
Two easy mistakes
First: when 3.7 shipped, Google moved 3.6 Flash onto the same intro rate. “3.7 is cheaper” is false; the gap is capability. Second: output price includes thinking tokens. High thinking hits the output line. Third-party “50% off” posts often mix Vertex promos with this table. We use the Developer API table only.
Price timeline
3.7 Flash GA. Intro rates live. 3.6 Flash moves to the same Standard intro.
Last day of intro pricing.
Standard $1.50 / $7.50, cache $0.15, Priority $2.70 / $13.50.
Confirmed vs treat carefully
Confirmed
Id gemini-3.7-flash. Standard intro $0.75 / $3.75 through 31 Dec 2026. Batch/Flex half. Priority $1.35 / $6.75. Cache $0.075 + $0.50/million/hour storage. From Google’s pricing page.
Treat carefully
Do not treat OpenRouter or Vertex promos as the official table. Do not compare today’s 3.7 to 3.6’s old $1.50 / $7.50 launch rate. All figures on this page come from Google's official table; QCode does not currently offer the Gemini series.
Who to compare against
Stay on 3.6
Intro list price is the same. Stay if the workflow is only validated on 3.6.
Budget for 3.7
Pay for the capability jump, and model 2027 as a double — do not annualize the intro rate.
How to estimate a call
Input at $0.75/million, output (including thinking) at $3.75/million. Cacheable prefixes at $0.075. Long agent loops are output- and thinking-heavy. Batch if you can wait.
On QCode
QCode does not currently offer the Gemini series; please use Claude, GPT or Chinese models (e.g. claude-sonnet-5-5, gpt-6.1-sol, glm-5.3). The figures on this page are Google's official price table, not QCode prices.
3.7 Flash pricing FAQ
What is the price today?
Official Standard: $0.75 in / $3.75 out per million tokens through 31 Dec 2026.
After the intro window?
$1.50 / $7.50 from 1 Jan 2027.
Is 3.7 cheaper than 3.6 Flash?
Not on the official Standard intro. Same list. The difference is capability.
Is thinking billed extra?
No separate line. Thinking tokens count as output.
What about Batch?
Intro $0.375 / $1.875. Flex matches. Priority is $1.35 / $6.75.
Can I call this rate on QCode?
No. QCode does not currently offer the Gemini series; please use Claude, GPT or Chinese models (e.g. claude-sonnet-5-5, gpt-6.1-sol, glm-5.3). The prices on this page are Google's official table.
Sources
Google Gemini Developer API pricing tables for gemini-3.7-flash and gemini-3.6-flash; Keyword 13 Aug 2026 intro footnote.
Price Gemini from Google's table; on QCode, use Claude, GPT or Chinese models
QCode does not currently offer the Gemini series; please use Claude, GPT or Chinese models. Google's intro price has an end date.
Related
Claude Code weekly limit, October 2026
No new official change found as of 2026-10-08: 5-hour sessions, fixed weekly reset time and your options after hitting the limit.
Claude Code on a third-party API: feature availability
As of 2026-10-08: local CLI, MCP, subagents, auto and 1M work; Remote Control, cloud sessions and other account-side features don't.
QCode pricing
How official list × rate works.
Not affiliated with Google. Figures are from Google’s pricing page; other channels may differ.