Gemini API Prices 2026 — Per-Model Costs, Examples, and Cheaper OpenAI-Compatible Access

Gemini is among the best-value leading AI models, and Kunavo offers it around 70% below Google's official list price through a single OpenAI-compatible API. This guide covers current rates by model, cost examples you can verify yourself, and the cheapest way to call Gemini in production.

Gemini API pricing at a glance

Rates are per 1M tokens, in USD, as billed by Kunavo. The “Google list” column shows Google's published official rates for the same models, so you can see the difference.

Flash is the workhorse for high volume; Pro is for more demanding reasoning, vision, and long-context tasks. Live rates are always listed on the pricing page and on each model page (gemini-2-5-flash, gemini-3-1-pro).

How Gemini pricing per token works

You pay for input tokens (everything you send — system prompts, retrieved context, user messages) and output tokens (what the model generates). Output is the more expensive side, so the biggest lever on your Gemini bill is how much text you let the model write. Images and audio are converted into token equivalents and billed on the same meter.

Cost calculation examples

Real figures using Kunavo's Gemini 2.5 Flash rates, except for the last row, which uses Gemini 3.1 Pro:

At those rates, classifying 100,000 documents in a Flash batch costs around $6, and one million chatbot turns cost around $315. The calculations are ready to run:

# Tarif Kunavo untuk Gemini 2.5 Flash (USD per 1J token)
IN_RATE, OUT_RATE = 0.09, 0.75

def biaya(token_masuk: int, token_keluar: int) -> float:
    return token_masuk / 1_000_000 * IN_RATE + token_keluar / 1_000_000 * OUT_RATE

print(biaya(1_000, 300))            # satu putaran chatbot      -> $0.000315
print(biaya(8_000, 500))            # satu jawaban RAG          -> $0.001095
print(biaya(500, 20) * 100_000)     # batch 100rb dokumen       -> ~$6.00

Kunavo pricing and Stripe billing

There are no subscriptions and no Google Cloud projects. Add funds to your wallet (via Stripe or a local payment method), and each call deducts from that balance at the per-token rates above. Pay-as-you-go starts with a $10 minimum top-up, balances never expire, and larger top-ups earn bonus credits. One wallet covers Gemini and all other models — Claude, GPT, image, video, and audio — so you don't have to reconcile separate bills from each provider.

Which Gemini model should I choose?

• gemini-2-5-flash — the default for chat, extraction, classification, summarization, and most RAG. Fast and the most capable low-cost option.
• gemini-3-1-pro — rely on it when Flash isn't accurate enough: multi-step reasoning, code, vision, and very long contexts.

A good pattern is to route by difficulty: Flash for common cases, moving up to Pro only when a check fails. See the AI cost optimization guide for routing patterns in code.

How to reduce your Gemini bill

1. Drop down a tier. Send 80% of easy cases to Flash; reserve Pro for the 20% that are difficult.
2. Limit outputs. Set max_tokens and stop sequences — output is the expensive side of the meter.
3. Trim inputs. Retrieve fewer, better RAG chunks instead of cramming your entire knowledge base into the context.
4. Batch. Combine independent calls to keep latency low and avoid retry storms.

Frequently asked questions