For the Gemini API, these two preview models are no longer an active head-to-head choice: Google lists gemini-3-pro-preview as shut down on March 9, 2026. Its lifecycle page lists gemini-3-flash-preview with no shutdown date announced, but names Gemini 3.6 Flash as its recommended replacement. This status check is dated November 9, 2026.
For a new Gemini API integration, don’t treat Gemini 3 Flash Preview as an unqualified current recommendation just because it has no shutdown date. Check the live model lifecycle and pricing pages before choosing an endpoint. The comparison below is about the Gemini API, not Google Cloud Agent Platform or consumer Gemini plans.
Gemini 3 Flash vs Pro: Gemini API comparison
| Detail | Gemini 3 Flash Preview | Gemini 3 Pro Preview |
|---|---|---|
| Gemini API model ID | gemini-3-flash-preview |
gemini-3-pro-preview |
| Lifecycle, checked Nov. 9, 2026 | No shutdown date announced; Gemini 3.6 Flash is listed as the recommended replacement | Shut down March 9, 2026 |
| Gemini API limits documented in the sources checked | 1,048,576 input tokens; 65,536 output tokens | No current usable endpoint; this comparison does not establish Pro’s limits |
| Original launch pricing | $0.50 per 1 million text input tokens; $3 per 1 million output tokens | $2 per 1 million input tokens and $12 per 1 million output tokens for prompts up to 200,000 tokens |
The lifecycle entries come from Google’s Gemini API deprecations page; Flash’s limits and capabilities come from its Gemini API model page.
What Gemini API documents for Flash
Google’s Gemini API model page lists Flash Preview inputs as text, image, video, audio, and PDF, with text output. It documents a 1,048,576-token input limit and a 65,536-token output limit. The reported Gemini API model-page support entries include caching, code execution, computer use, and file search. The page also lists Batch API, Flex inference, and Priority inference as consumption options. Confirm each required feature against the live documentation for the specific API request path you plan to use; the cited support entries do not establish every feature’s request-path restrictions.
The model page identifies this as a preview model. The lifecycle page currently lists no Flash shutdown date, but recommends Gemini 3.6 Flash as its replacement. Those are separate facts: “no shutdown date announced” is not the same as a recommendation to start new production work on the preview endpoint.
Historical API pricing—not a current cost estimate
Google’s December 17, 2025 launch announcement listed Gemini 3 Flash at $0.50 per 1 million input tokens and $3 per 1 million output tokens through the Gemini API and Vertex AI. It noted a different input rate for audio: $1 per 1 million audio input tokens. Google’s Gemini 3 developer announcement listed Pro Preview at $2 per 1 million input tokens and $12 per 1 million output tokens for prompts of 200,000 tokens or less. These are launch prices, not verified current rates.
At those historical rates, a hypothetical workload of 100,000 text input tokens and 10,000 output tokens would calculate to $0.08 for Flash and $0.32 for Pro. This arithmetic excludes other possible charges and does not represent a present-day deployment estimate. In particular, the Pro Gemini API endpoint is shut down, so its launch price is not a usable current alternative. Check Google’s current pricing page for any model you can actually call.
Which should you choose?
- Comparing these exact Gemini API endpoints: Pro Preview is shut down. Flash Preview has no shutdown date announced, but Google lists Gemini 3.6 Flash as the recommended replacement.
- Checking whether Flash fits a workload: use the Gemini API model page’s input, output, tool, and capability details as a compatibility checklist. Confirm that each required feature works through the API method you plan to use.
- Budgeting: don’t use the historical launch rates above as current pricing. Verify live rates and applicable billing conditions for the model and API available to your project.
- Choosing on quality or speed: the sources cited here do not establish a controlled, current Flash-versus-Pro performance comparison. Pro’s Gemini API endpoint is retired, so a like-for-like live API test between these exact endpoints is not possible.
Bottom line
For the Gemini API, Gemini 3 Pro Preview shut down on March 9, 2026. Gemini 3 Flash Preview has no shutdown date announced in Google’s listed lifecycle entry, but the same entry recommends Gemini 3.6 Flash as its replacement. Flash’s API documentation specifies its modalities, supported capabilities, and token limits; its original launch prices are historical, not a current cost verdict.
Sources
Gemini deprecations | Gemini API
Gemini 3 Flash Preview | Gemini API
Build with Gemini 3 Flash, frontier intelligence that scales with you
Gemini 3 for developers: New reasoning, agentic capabilities