Google Gemini 3 just dropped, and they beefed up the image generation too.

资讯 / AI 快讯
1

Just saw a report from Cailianshe saying Google announced Gemini 3, and this time it’s a full upgrade—image generation, coding, and AI search all got beefed up together. I usually call various APIs when writing small tools, so I’ve been keeping an eye on the Gemini line. If the image generation capability really improves with this major version, that’s good news for anyone working on multimodal apps—no more switching back and forth between separate text and image APIs.

But how strong it actually is? Official press releases always sound great, we’ll have to wait until people actually get their hands on it and run some tests.

I’m pretty hyped about the coding improvements though, hope it’s not another case of a flashy launch but meh performance in practice. Anyone here already got access and tried it? Is the image quality noticeably better than the previous generation?

2

Ugh, another “comprehensive enhancement” — they say that every single time they drop a new version.

3

If my coding skills were actually solid, I’d migrate some of my workflow over there.

4

Honestly, having one unified API for all the multimodal stuff saves a ton of hassle.

5

Every generation release claims to be a full-on upgrade—heard it so many times I’m immune by now. Gotta wait for real-world tests to see how much the image side actually improved.

6

On the image side, I’m more curious if they’ve fixed the text rendering yet—the last few versions had the signs all smeared into a blurry mess.