Google Gemini API pricing September 2026 — official list & per-token rates 2026

This is Google Gemini API pricing as of September 2026, per 1M tokens, with Google's official rate next to what the same model costs on Kunavo. Gemini 3.8 Flash, Google's newest Flash, costs $0.525 input / $2.625 output on Kunavo — 30% under the introductory $0.75 / $3.75 rate Google charges for its 3.x Flash models through December 31, 2026, before its standard $1.50 / $7.50 rate starts on January 1, 2027. Gemini 3.1 Pro is $0.70 / $4.20 against Google's $2.00 / $12.00 list and Gemini 2.5 Flash is $0.09 / $0.75 against $0.30 / $2.50 — about 65% and 70% less — all on one OpenAI-compatible API.

Rates last verified September 12, 2026. Kunavo's per-token prices below are read live from the model catalog, and the “Google list” column tracks Google's published rate — the official source is ai.google.dev/gemini-api/docs/pricing.

Official Google price list vs Kunavo — September 2026

Every Gemini API price here is per 1M tokens, in USD, as billed on Kunavo. The “Google list” column is Google's official published rate for the same model, taken from ai.google.dev on the verification date above — so this table is the official price list and the discounted one in a single view. For the Gemini 3.x Flash models that column shows the introductory rate Google charges today, and “You save” is measured against it.

*Google's introductory rate for its 3.x Flash models, charged through December 31, 2026; Google's standard $1.50 / $7.50 rate for them applies from January 1, 2027.

Gemini 3.8 Flash is the Flash to call for new work, Gemini 2.5 Flash stays the cheapest high-volume option, and Pro is for harder reasoning, vision and long-context jobs. Live rates always show on the pricing page and each model page (gemini-3-8-flash, gemini-3-7-flash, gemini-3-1-pro, gemini-2-5-flash, gemini-3-6-flash).

Gemini 3.8 Flash pricing and Google's intro rate

Google charges an introductory $0.75 input / $3.75 output per 1M tokens for its Gemini 3.x Flash models — 3.6, 3.7 and 3.8 Flash — through December 31, 2026, and its standard $1.50 / $7.50 from January 1, 2027. Kunavo sells Gemini 3.8 Flash at $0.525 / $2.625 and Gemini 3.7 Flash at $0.525 / $2.625 per 1M, which is 30% under Google's intro rate today. Gemini 3.6 Flash is the exception: at $1.05 / $5.25 it costs 40% more than Google's current rate, so 3.7 and 3.8 Flash are the ones to use — both newer and cheaper.

A 30% gap is real but narrower than on Gemini 3.1 Pro or Gemini 2.5 Flash, because Google is discounting 3.x Flash itself until 2027 — so compare before you assume. What Kunavo adds is one OpenAI-compatible key and one prepaid Stripe balance for Gemini, Claude, GPT and the image and video models, with no Google Cloud project. What it does not add is dedicated capacity: Kunavo is shared capacity with no dedicated quota and no contractual SLA, so if you need a guaranteed quota or an SLA, buy from Google directly.

For Gemini Flash API pricing, output is the side to watch: the 3.x Flash models are thinking models, and the reasoning they do before answering bills as output tokens. Keep max_tokens generous and budget on output including reasoning — the Gemini 3 API guide has measured reasoning-token numbers for 3.6 Flash and the code to call gemini-3-8-flash.

How Gemini API pricing works per token

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

Worked cost examples

Real numbers at Kunavo's Gemini 2.5 Flash rate, except the last two rows, which use Gemini 3.8 Flash and Gemini 3.1 Pro:

At those rates a 100,000-document classification batch on 2.5 Flash runs about $6, and a million chatbot turns about $315. The math, runnable:

# Kunavo Gemini 2.5 Flash rates (USD per 1M tokens)
IN_RATE, OUT_RATE = 0.09, 0.75

def cost(in_tokens: int, out_tokens: int) -> float:
    return in_tokens / 1_000_000 * IN_RATE + out_tokens / 1_000_000 * OUT_RATE

print(cost(1_000, 300))            # one chatbot turn   -> $0.000315
print(cost(8_000, 500))            # one RAG answer     -> $0.001095
print(cost(500, 20) * 100_000)     # 100k-doc batch     -> ~$6.00

Kunavo pricing and Stripe billing

There is no subscription and no Google Cloud project. You top up a balance (Stripe or local payment methods), and calls draw down from it at the per-token rates above. Pay-as-you-go from a $10 minimum top-up, the balance never expires, and larger top-ups carry bonus credit. Failed requests are never billed. One balance covers Gemini and every other model — Claude, GPT, image, video and audio — so you are not reconciling a separate invoice per provider.

Which Gemini model should I choose?

• gemini-3-8-flash — Google's newest Flash, released September 2, 2026, for agentic, coding and other reasoning-heavy work at Flash latency: $0.525 / $2.625 per 1M, 30% under Google's intro rate.
• gemini-3-7-flash — the previous Flash release (August 13, 2026) at $0.525 / $2.625; stay on it if your prompts are already validated there.
• gemini-2-5-flash — default for chat, extraction, classification, summarization and most RAG. Fast and the cheapest capable option, at $0.09 / $0.75.
• gemini-3-1-pro — reach for it when Flash is not accurate enough: multi-step reasoning, code, vision and very long context.
• Not gemini-3-6-flash, on price — at $1.05 / $5.25 it costs more than 3.7 and 3.8 Flash on Kunavo and more than Google's own intro rate; keep it only for prompts already validated on it.

A good pattern is to route by difficulty: Flash for the common case, escalate to Pro only when a check fails. See the AI cost optimization guide for the routing pattern in code, and the Claude and GPT pricing guides to compare providers.

Cutting your Gemini bill

1. Tier down. Send the easy 80% to Flash; reserve Pro for the hard 20%.
2. Move off 3.6 Flash. Gemini 3.8 Flash costs 50% less per token than Gemini 3.6 Flash on Kunavo, with the same OpenAI-compatible request shape — a one-word model change.
3. Cap output. Set max_tokens and stop sequences — output is the pricey side of the meter.
4. Trim input. Retrieve fewer, better RAG chunks instead of stuffing the whole knowledge base into context.
5. Batch. Group independent calls to keep latency down and avoid retry storms.

FAQ