Gemini 3 Flash (preview) pricing
Google's newest Flash preview. 1M context window at bargain rates - no long-context surcharge to worry about.
Input
$0.50/ 1M tok
Output
$3.00/ 1M tok
- Context window
- 1M
- Max output
- -
- Cached input
- $0.050 / 1M
- Verified
- 2026-04-06
Gemini 3 Flash (preview) is the Flash tier of Google's next-generation line. At $0.50 per 1M input and $3 per 1M output it's 4x cheaper than Gemini 3.1 Pro (preview) on both sides, and unlike the Pro tier, there's no long-context surcharge - pricing stays flat across the full 1M context window.
Cached input at $0.05 per 1M gives a 90% discount on repeated prefixes. For bulk workloads like document ingestion, summarisation, and classification where the system prompt or reference material repeats, setting up Google's context caching is usually worth the wiring.
Calcis counts input tokens against Google's own countTokens API so the number matches what Google uses to bill, and predicts response length from the prompt so you can see a full dollar estimate before you send.
Estimate your cost on Gemini 3 Flash (preview)
Paste your prompt into the estimator, pick Gemini 3 Flash (preview), and see the exact dollar cost - input tokens counted with the provider's own tokenizer, output tokens predicted by our regression model.
Frequently asked
- Does Gemini 3 Flash have a long-context surcharge?
- No. The $0.50 / $3 per 1M rates apply across the full 1M context window. Only the Pro tier (3.1 Pro and 2.5 Pro) has dual-tier pricing above 200K input tokens.
- How much does Gemini 3 Flash cost per request?
- A 1,000-token prompt with a 500-token reply costs about $0.002 ($0.0005 input + $0.0015 output). At 1 million requests a month, that's around $2,000 - very cheap for a frontier-class model.
- Is Gemini 3 Flash cheaper than Gemini 2.5 Flash?
- No - 2.5 Flash is cheaper at $0.30 / $2.50 per 1M tokens. 3 Flash is a newer preview with different capabilities; pick the preview if you need the newer generation, stay on 2.5 Flash for cost-optimised GA workloads.
- What tokenizer does Gemini 3 Flash use?
- Google uses SentencePiece but doesn't publish it for offline use. Calcis calls Google's countTokens API for the authoritative count, which matches the Google billing boundary exactly.
Pricing verified 2026-04-06 from the provider's rate card.