Gemini

Gemini

gemini.google.com
Google's multimodal AI assistant, built with Google DeepMind
Full descriptionGemini is an interface to a multimodal model that handles text, audio, images and more. Gemini in Pro has a context window of 1 million tokens, enough to process up to 1,500 pages of text or 30K lines of code at once. It can summarize long documents, help with coding, debug errors, and generate images and video.
Best for
General chat and Q&A
Best for

Reading through long documents and long material

Watch out

Same question, different answers; verify facts

Visit site
✓ Recommended 5◐ Good enough 4✕ Failed 4
Price
Free to use
Access
Not accessible from mainland China

User verdicts

User verdicts · independent of our own tests above
38%
Recommend rate · 13 verdicts
How was it for you?
·
Huiyong AI userFailedJun 20, 2026

更新完就变了个样，之前能记住的上下文现在记不住了，回复也短了一截。以前那种能陪着做长项目的感觉没了。

Original in Chinese
NicoRecommendedJun 11, 2026

两家我都订着，结果发现自己开它的次数明显更多，想点子、理数据、看资料基本都在这边。正琢磨要不要把另一个停掉。

Original in Chinese
Huiyong AI userFailedJul 24, 2026
See all verdicts >

Alternatives

Interchangeable picks for the same job · each with its own evidence
ChatGPTOpenAI's free all-round AI assistantNot tested yet · info from official sources
元宝All-round, not a standout: fine for search, writing and summaries; for ready-to-edit files, add a specialist tool.Compared in testsJul 7
豆包Chat, Q&A, creative content generation and moreNot tested yet · info from official sources

Know a better AI tool?

Recommend a tool you like. It takes seconds, and it goes live after review.

Latest updates

Original in Chinese
2026 / 104 updates
14
VersionGoogle 发布新前沿模型 Gemini 4 Argon，并正在 Gemini 应用中逐步推送。
Via X：Gemini (@GeminiApp) · Read original ↗
14
VersionGoogle DeepMind 的 Gemini 4 Argon 模型上线 OpenRouter，可通过 Google AI Studio 和 Google Cloud 调用。
Via X：OpenRouter (@OpenRouter) · Read original ↗
06
VersionGoogle 发布 Gemini Nano Banana 2.1 图像生成与对话式编辑模型，旧版 gemini-3.1-flash-image 将于 2026 年 10 月 29 日停用。
Via Gemini API：更新日志（网页） · Read original ↗
01
VersionGoogle DeepMind 发布新前沿模型 Gemini 4 Argon，先通过 Fairwind Program 向可信网络防御者开放，后续再面向开发者、企业和消费者推出。
Via Google DeepMind：Blog（RSS） · Read original ↗
2026 / 094 updates
23
VersionGoogle DeepMind 发布 Gemini 3.8 Flash TTS 与 Flash-Lite TTS 两款文本转语音模型，支持用提示词设计声音、30 秒样本复刻和双说话人编排，覆盖 100 多种语言。
Via Google DeepMind：Blog（RSS） · Read original ↗
16
VersionGoogle DeepMind 发布 Gemini 3.8 Live 和 Gemini 3.8 Live Extended Thinking 两个近实时语音对话模型。
Via Google DeepMind：Blog（RSS） · Read original ↗
03
VersionGoogle DeepMind 发布 Gemini 3.8 Flash 与 3.8 Flash Cyber 两款新模型。
Via Google DeepMind：Blog（RSS） · Read original ↗
02
New featureGoogle DeepMind 为 Gemini 3.7 Flash、3.6 Flash 和 3.5 Flash-Lite 推出 agentic 视频理解功能，可动态扫描视频片段。
Via Google DeepMind：Blog（RSS） · Read original ↗
RecordAI searchKnowledge-base Q&ALong-form writingWeChat articlesSpotted an error?

我们使用 first-party Cookie 做匿名访问统计，以改进内容与体验。匿名标识不关联个人身份信息（手机号/邮箱）， 详见《用户协议及隐私政策》。