Gemini API Cost Calculator

By the TokenForge team · Last updated July 2026

What will the Gemini API cost you per month? This TokenForge preset covers Google's lineup — Gemini 3.5 Flash, Gemini 3.1 Pro and the budget 2.5 Flash-Lite — a range where the cheapest tier is nearly 20× cheaper than the mainstream one, so tier choice decides the budget. Enter your volume and ratio; prices refresh daily.

Gemini API cost tool

M
: 1
%
Ops overhead
Throughput presets (published benchmarks)

At 50M tokens/month, the API is cheaper by $1,959.70/month.

API cost $150.00/mo Gemini 3.5 Flash · $3.00 per 1M blended
Self-host cost $2,109.70/mo RunPod · H100 80GB · $0.90 per 1M at capacity
Break-even volume 703M tokens/mo Server capacity ≈ 2.33B tokens/mo

† Anthropic's newest models (Fable 5, Opus 4.8, Sonnet 5) tokenize the same text into ~30% more tokens; the calculator corrects for this. Prices exclude batch (−50%) and prompt-caching discounts.

Prices updated July 10, 2026

What Gemini costs per month (July 2026)

Gemini modelInput $/1MOutput $/1MBlended @4:150M/mo500M/mo
Gemini 3.1 Pro (≤200k ctx)$2.00$12.00$4.00$200$2,000
Gemini 3.5 Flash$1.50$9.00$3.00$150$1,500
Gemini 2.5 Flash-Lite$0.10$0.40$0.16$8$80

Monthly figures assume a 4:1 input:output ratio. The step from Flash to Flash-Lite is the striking one: an 18.75× price gap between two tiers of the same family, which makes "route the easy queries down" the single most effective Gemini cost strategy.

How the bill is calculated

Worked example. A RAG search assistant serving 50,000 queries a day at roughly 1,600 input and 400 output tokens each uses about 3B tokens/month at 4:1. On Gemini 2.5 Flash-Lite that costs 3,000 × $0.16 = $480/month. The same traffic on Gemini 3.5 Flash is $9,000, and on 3.1 Pro $12,000 — routing easy queries to the cheap tier is where Gemini budgets are won.

Picking a Gemini tier — context windows and caching

Three tiers with clean roles. Gemini 3.1 Pro is the reasoning tier for the hardest work; Gemini 3.5 Flash is the mainstream default with the best quality-per-dollar for user-facing features; and 2.5 Flash-Lite at $0.16 blended handles classification, extraction, routing and other high-volume chores for pocket change. Because all three sit behind the same API, tier-routing is a config change, not a rewrite.

Watch the context-window pricing on Pro. The 3.1 Pro price in this calculator is Google's standard rate for prompts up to 200k tokens of context. Prompts beyond that bill at higher per-token rates, so long-context workloads — full-codebase analysis, book-length documents — should budget above the number shown here. If your prompts are that large, your input:output ratio is also far above 4:1; raise it in the tool to see the real blended price.

Caching and batch cut deep. Context caching makes repeated input tokens about 90% cheaper ($0.15 vs $1.50 per 1M on 3.5 Flash) — valuable for assistants that hold a large document or system prompt across turns. Batch mode halves prices for asynchronous jobs. And for prototyping, Google AI Studio's rate-limited free tier means the meter only really starts when you ship. As always, the calculator shows undiscounted list prices — your ceiling, not your floor.

Self-hosting rarely beats Gemini's budget tiers. A rented H100 at the preset assumptions overtakes Gemini 3.5 Flash from about 703M tokens/month, but against Flash-Lite the break-even (~13B tokens/month) is far beyond a single server's capacity. If you're weighing that trade, start by estimating your real monthly token usage, then compare every provider in the main calculator.

Frequently asked questions

Related tools and guides