| Spec | Value |
|---|---|
| Maker | Google DeepMind |
| Status | Preview (gemini-2.5-flash-preview-04-17) |
| Thinking | Hybrid: thinking on or off, with an adjustable budget |
| API price per 1M tokens (input / output) | $0.15 / $0.60 (no thinking) or $3.50 (thinking) |
From Google's launch benchmark table image (prices as of April 17, 2025; single-attempt scores); the launch posts state no context window or maximum output. · source
Estimate from the version's published specs, benchmarks and reviews, 0-5; scores are only comparable within one criteria year. Scored as launched (Gemini 2.5 Flash preview, 17 Apr 2025). Reasoning null: AA index at launch not on the 2026 scale (AA's page now shows 9 non-thinking / 12 thinking on its rescaled index); Google published no headline score comparable to the anchors. Coding null. Context 1M tokens -> 4 (9to5Google; Google blog). Price 5: Google pricing page (Wayback 18 Apr 2025) $0.15 in / $0.60 out non-thinking (about $0.26 blended) or $3.50 out with thinking (about $0.99 blended); both score 5. Speed null. Modalities 5: text, images, video, audio in. Reviews: Simon Willison (hands-on, the 17 Jun GA post also tests the GA Flash), 9to5Google launch report; independent hands-on for the April preview is thin.[1]
The way the model is priced is a little complicated.
I got good results from 2.5 Pro (74,073 input, 8,856 output = 18.1151 cents , 147.5 seconds) and from 2.5 Flash (74,073 input audio, 10,477 output = 10.026 cents , 72.6 seconds)
If the thinking budget is set to zero, this new model will match 2.0 Flash’s cost & latency.
| Start building with Gemini 2.5 Flash · Apr 17, 2025 | -0.20 |
|---|---|
| Gemini 2.5 Flash with ‘thinking budget’ rolling out to devs, Gemini app · Apr 17, 2025 | 0.20 |
| Trying out the new Gemini 2.5 model family · Jun 17, 2025 | 0.40 |
Photos from readers, kept apart from the pin's sourced media. Rate what you ate or saw, and on a restaurant pin tag the dishes right on your photo.
The first entry is always the pin's source. Overall confidence is a weighted average of how firmly each reference supports the start and end times used above; a reference counts half as much for every 180 days older than the newest.
The Google Developers Blog[3] post "Start building with Gemini 2.5 Flash" is dated 17 April 2025 and says the preview is available that day in the Gemini API via Google AI Studio and Vertex AI.
The Wikipedia article lists Gemini 2.5 Flash as released on 17 April 2025.
Google's 17 June 2025 post made 2.5 Flash and Pro generally available and added 2.5 Flash-Lite.[1]
Something missing or wrong? Say it in your own words: a link that backs this pin up, a different start or end date and why, or a fact it lacks or gets wrong. The AI checks it against this pin's sources, searches for better ones, and adds any page that backs you up. The pin's own sources still count most. A picture that shows something else, or shows it badly, is looked at too, and moved down or replaced.