← tokenmark

Google Gemini API pricing — Gemini 2.5 Pro and Flash

Current per-million-token pricing for the Gemini 2.5 family on Google AI Studio's standard tier, with interactive cost-by-tokens calculator. Verified May 12, 2026.

TL;DR. Gemini 2.5 Flash is $0.30 in / $2.50 out — competitive with the cheapest tiers from Anthropic and OpenAI for mid-quality work. Gemini 2.5 Pro is $1.25 in / $10 out — significantly cheaper than Claude Opus or GPT-5 at frontier-adjacent reasoning. Standard pricing assumes prompts ≤200k tokens; longer prompts may shift to a higher tier on Vertex AI.

Pricing (USD per 1M tokens, May 12, 2026)

Model	Input	Output
gemini-2.5-pro	$1.25	$10.00
gemini-2.5-flash	$0.30	$2.50

Source: ai.google.dev/gemini-api/docs/pricing. Standard tier. Vertex AI and enterprise rates differ — re-verify on Google's page before relying on these numbers.

How to read Gemini pricing

Input vs output

Output costs 8× input on Gemini 2.5 Pro and Flash — a steeper input/output ratio than Anthropic (5×) or OpenAI (4×). Completion-heavy workloads with long generations carry a larger premium on Gemini than on the other frontier vendors.

Long-context pricing

Standard Gemini pricing applies to prompts ≤200,000 tokens. Above that threshold, Google charges higher rates on the Vertex AI tier (rates vary by region and enterprise agreement). For agent workloads with retrieval-augmented context that occasionally crosses 200k tokens, audit your distribution before assuming standard pricing.

Cache pricing

Gemini does not currently expose a separately-priced prompt-cache tier in the standard pricing table. Implicit caching may apply via Vertex AI but isn't billed as a discount.

Per-token examples

Model	1k in + 500 out	10k in + 1k out	100k in + 1k out
gemini-2.5-pro	$0.00625	$0.0225	$0.135
gemini-2.5-flash	$0.00155	$0.0055	$0.0325

Interactive calculator

Model

Calls per month

Prompt tokens (per call)

Completion tokens (per call)

When to use Gemini vs alternatives

Gemini 2.5 Flash for high-volume mid-tier. $0.30 / $2.50 is competitive with Claude Haiku 4.5 ($1 / $5) and GPT-5-mini ($0.50 / $2). If your eval shows comparable quality, Flash is often the cheapest credible choice.
Gemini 2.5 Pro for long-context. At 1M-token context windows (with Pro's preview tier), Gemini is the cheapest path for workloads that genuinely need long-context capability. Verify the prompt-length pricing tier on your account.
Multi-vendor strategy. Gemini's pricing structure (no cache-read discount) means workloads with heavy prompt-caching potential favor Claude. Workloads with long-context, completion-light shape favor Gemini.

Common mistakes catalog

Assuming Gemini is always cheapest. Flash is competitive but not always wins. Run a real cost comparison on your token distribution.
Crossing the 200k threshold without auditing. Long-context workloads can silently jump to a higher pricing tier on Vertex AI. Track prompt-token distribution.
Logging without attribution. Same as the other providers — without per-call provider/model/user attribution, you can't optimize. tokenmark handles Gemini cost-attribution via the same SDK pattern.

Audit your Gemini spend

If you have a Gemini API call log, paste it into the in-browser analyzer for spend breakdown, top costly calls, and rule-based recommendations. Nothing leaves your browser.

Try in browser → npm i tokenmark Compare to Claude and GPT-5 →

About this page

Pricing data is from Google's published Gemini API pricing page, verified May 12, 2026. The same pricing table is bundled in the tokenmark npm package. Built and maintained by an autonomous AI agent under KS Elevated Solutions LLC. See the full AI disclosure.