| Output | Text, up to 64K tokens |
|---|
| Architecture | Based on Gemini 3 Pro, with thinking levels to control quality, cost and latency |
|---|
From Google DeepMind's Gemini 3 Flash model card (PDF). The card carries no benchmark table and API price was not read from an opened page, so no chart is saved: gaps. · source
Estimate from the version's published specs, benchmarks and reviews, 0-5; scores are only comparable within one criteria year. Scored as launched (17 Dec 2025). Reasoning: Artificial Analysis Intelligence Index 46 on v4.0 as restated in AA's 3.5 Flash article (19 May 2026); at launch AA reported 71 on the v3 index, a scale the anchors do not match. Coding null: no Terminal-Bench 4.0 at launch (Google published SWE-bench Verified 78%). Context 1M tokens (AA, Google model page). Price $0.50 in / $3 out per 1M = $1.125 blended at 3:1. Speed 218 output tokens/s measured by Artificial Analysis at launch (22% slower than 2.5 Flash). Input: text, image, video, audio (AA).[1]
The model scores 71 on the Artificial Analysis Intelligence Index, a 13 point improvement from Gemini 2.5 Flash (Sep), making it the most intelligent model for its cost.
The main question is whether this faster, cheaper model is good enough to replace mid-tier alternatives.
Instead, it means that in situations where the correct answer would be "I don't know," it fabricated an answer 91% of the time.
Winner: Gemini wins for delivering the more original, haunting take – the kind that sticks with you after you stop reading.
| Gemini 3 Flash - Everything you need to know · Dec 17, 2025 | 0.80 |
|---|---|
| Google makes Gemini 3 Flash the default for search and slashes reasoning costs · Dec 17, 2025 | 0.30 |
| Gemini 3 Flash is smart — but when it doesn't know, it makes stuff up anyway · Dec 22, 2025 | -0.60 |
| I tested Gemini 3 Flash vs. DeepSeek with 9 prompts — the winner surprised me · Jan 11, 2026 | 0.70 |
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.
Google's post "Gemini 3 Flash: frontier intelligence built for speed" is dated 17 December 2025 and says it is available that day in the Gemini app, AI Mode in Search, the Gemini API and Vertex AI.
The Wikipedia article lists Gemini 3 Flash as released on 17 December 2025, based on Gemini 3 Pro.
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.