The Gemini 3.x rate table
Google prices the Gemini family per million tokens like everyone else, with the Flash line covering the budget-to-mid tiers and Pro at the top. The standout item: Gemini 3.8 Flash's $0.75/$3.75 rate is an introductory price through December 31, 2026, stepping up to $1.50/$7.50 in January 2027, per Google's API pricing (ai.google.dev) — verified October 28, 2026.
| Model | Input / 1M | Output / 1M | Pricing source |
|---|---|---|---|
| Gemini 3.8 Flash$1.50/$7.50 from Jan 1, 2027 | $0.75 | $3.75 | Google API pricing (ai.google.dev) |
| Gemini 3.1 ProPrompts >200K tokens: $4/$18 | $2.00 | $12.00 | Google API pricing (ai.google.dev) |
| Gemini 3.5 Flash-Lite | $0.30 | $2.50 | Google API pricing (ai.google.dev) |
| Gemini 3.1 Flash-Lite | $0.25 | $1.50 | Google API pricing (ai.google.dev) |
| Gemini 3 Flash Preview | $0.50 | $3.00 | Google API pricing (ai.google.dev) |
| Gemini 2.5 Flash-LiteCheapest Gemini entry point | $0.10 | $0.40 | Google API pricing (ai.google.dev) |
Source transparency: every row in this table is taken from Google's own API pricing page (ai.google.dev), retrieved October 28, 2026. An earlier edition of this page sourced some rows from third-party pricing indexes; those rows have since been re-verified against Google's page or removed. See our dashboard for the full verification policy.
How Gemini compares on price
At intro pricing, 3.8 Flash ($0.75/$3.75) undercuts GPT-6.1 Sol ($2/$10) on input by more than half while sitting above DeepSeek V4.1 Flash ($0.15/$0.60 off-peak). At 2027 pricing ($1.50/$7.50), it lands between the Flash-Lite tiers and GPT-6.1 Sol. The Flash-Lite tiers ($0.10–$0.30 input) compete in the budget band with DeepSeek and GPT-6 Luna ($0.10/$0.50).
Two Google-specific billing notes from Google's pricing page: Gemini bills cache storage by the hour on top of cache-read discounts, and some Flash tiers carry a free tier whose content may be used to improve Google products — the standard paid-vs-free tradeoff.
Which Gemini tier for which job
- 3.8 Flash — the current default for general production work; lock 2027 budgets at the stepped-up rate
- 3.1 Pro — the capable tier for harder tasks; watch the >200K-token band ($4/$18)
- Flash-Lite tiers — high-volume simple work: classification, extraction, and the same bulk jobs DeepSeek and Luna compete for
Compare every tier side-by-side on the pricing dashboard, and model your own mix with the bill estimator.
Quick answers
How much does Gemini 3.8 Flash cost?
Verified October 28, 2026 on Google's API pricing page: $0.75 per million input tokens and $3.75 per million output tokens — an introductory rate through December 31, 2026. From January 1, 2027, the rate steps up to $1.50/$7.50, so budget 2027 work at the stepped-up rate, not the intro rate.
Is the Gemini API cheaper than OpenAI?
At intro pricing, Gemini 3.8 Flash ($0.75/$3.75) undercuts GPT-6.1 Sol ($2/$10) significantly on both input and output. After the January 2027 step-up to $1.50/$7.50, the gap narrows but does not close. DeepSeek V4.1 Flash remains cheaper than both at $0.15/$0.60 off-peak, before any quality comparison.
What is the cheapest Gemini model?
Gemini 2.5 Flash-Lite at $0.10/$0.40 per million tokens, verified on Google's API pricing page — the cheapest Gemini entry point. The current-generation budget pick is Gemini 3.8 Flash at $0.75/$3.75 intro pricing, and Gemini 3.5 Flash-Lite sits between them at $0.30/$2.50 for cost-efficient current-generation work.
Does Gemini charge for long context?
Yes, on the Pro tier: Gemini 3.1 Pro bills $4/$18 per million tokens for prompts above 200K tokens, against $2/$12 at or below 200K — and the higher band applies to the whole request. The Flash tiers shown here carry no long-context band on Google's current pricing page.