
Discover The Best AI Tools https://xitter.cf/t.co/YFo9GigY64 https://xitter.cf/t.co/IQXEukia8p Linktree link: https://xitter.cf/t.co/qNZNwfwtVW #AINavigation#toolnav
Joined June 2023
- Tweets1.6K
- Following1.5K
- Followers1K
- Likes817
X 合并订阅以后,就一定更省钱吗?
The X Subscription 把 X、Grok、Cursor 合并订阅了,包含 Grok Bot 的 Premium 档为 $30/月。
这个30刀真的划算吗?
如果本来就在为这三个产品付费,合并可能更划算。
但如果主力工具是 Codex 或 Claude,只想试试 Grok Bot,这 $30 就成了额外开支。
所以这次订阅 Grok Bot之前,我比较了 GamsGo @GamsGo_Global的方案。
我重点看的是 X Subscription Premium 独享套餐,订阅 1 个月,用我的优惠码 ANFZA 下单再省 3%,结算价是 US$23.76,比官方 US$30 的月付标价少 US$6.24,先试一个月,好的话再考虑买年费。(具体权益以套餐说明为准~)
GamsGo 上还有 ChatGPT、Gemini、Perplexity,以及 Netflix、YouTube、Office、Canva。整理账单时,可以顺便看看常用的几个软件有没有更划算的方案,留意独享、共享或个人账号充值的区别。
价格之外,我也会看交付和售后。平台有订单、客服入口,以及 GamsCare+ 保障选项,补发和退款条件可以提前查看,省得买完再来回找客服沟通。
最后提醒老用户:旧套餐可以保留,切换自选。用得顺手的先留着,有新增需求再比较,没必要每一项都原价续费。
toolnav.ai retweeted
Want a DR 36 backlink? Add your site to toolnav.ai
for free
toolnav.ai retweeted
元宇宙又复活了。
Meta 眼镜 这次把 VR 干到了 100 克。
比 Quest 3 轻差不多 5 倍,$1300,明年春天发货。
刚好我这两天在云栖大会逛了一圈。
一个非常强烈的感觉:
🔥做 AI 视频的公司,已经多到离谱了。
走几步一家。
广告视频、商品视频、数字人、短剧、自动剪辑……
大家底层接的模型其实也越来越像。
所以我现在看一个 AI 视频产品,已经不太关心:
你又接了几个模型。
能不能让我少操心才关键。
如果乔布斯还在,他会怎么给这副 Meta 眼镜拍一条广告?
🔥我直接 0 帧起手,嘴炮 Pexo。
要求就几个:
100g、轻、未来感、VR 真正变成一副眼镜。
然后让它自己出脚本、拆镜头、做关键帧,再一段段生成。
哪里不对继续聊。
Pexo还能调整细节,直接用 Mark to Fix 圈出来改。
这是 Pexo 跟我在展会上看到一堆视频工具,最大的区别。
现在这个行业太卷了:
Veo、Kling、Seedance、万相……模型一大堆。
15 秒、30 秒、不同分辨率、不同价格、不同积分……
普通用户哪有空天天研究这些破玩意。
一个真正有用的镜头,可能就 3~5 秒。
该用哪个模型,让 Agent 自己选。
该拆成几个镜头,让 Agent 自己想。
Motion、重生成、剪辑、音乐、字幕……
这些东西最好都别让我看到。
AI 视频公司越多,@Pexoai_offical 这种产品反而越有意义。
把这一堆 AI 视频工具,全藏到一个 Agent 后面。
何同学的 iPhone 18 Pro 宣传片夯爆了。
他这个视频到底用了什么模型?传统 CG 艺术家也太强了。
这种以前明显要导演 + 剪辑 + 动效 + CG 才能做的片子,现在普通人能不能拿 AI 搞?
刚好最近我在测一个 AI 视频产品 Pexo。
我直接 0 帧起手,嘴炮 Pexo,跟它边聊边把视频做出来。
🔥 开始我就提要求,我要一支运动鞋广告。镜头钻进鞋子内部,鞋底液态成型、鞋面现场织出来、鞋带自己穿孔。
几十只鞋在流水线上狂奔,最后一只直接“跑”到真人脚上,再收一个干净的产品特写。
先看鞋款设计图和关键帧,确认样子,再看“镜头钻进鞋里”这段转场。方向对了,继续往下做。
当然,镜头出来也不是直接交卷。
比如“鞋跑到脚上”这一段,中间有版鞋飞过去了,脚却没套进去。镜头再炫,鞋没穿上也得重做。
画面里有哪个地方想改,还能用 Mark to Fix,直接圈出来,在右边的批注框里写清楚要求,再提交修改。跟给文档留批注差不多。
最后做出来,是一条 31 秒的完整视频。
🔥以前我做 AI 视频特别喜欢干一件蠢事:上来先选 15 秒、30 秒,选最好最贵的模型。
现在回头看:全他妈浪费钱。😂
一个镜头真正有用的部分可能就 3~5 秒。
比如鞋底成型,4 秒已经结束了。
你偏要生成 15 秒,后面 11 秒模型只能开始自己发挥:
鞋开始变形、镜头开始乱飘。
最后你花了 15 秒的钱,真正剪进成片的可能只有 4 秒。
包括不同的生成镜头,也不是一个模型全部包圆。
Seedance、Kling、Hailuo 本来就各有所长。
Pexo爽点在,终于不用为了做一条视频,先把各家模型研究一遍了。
剩下的模型选择、镜头拆分、生成、Motion、剪辑、音乐、字幕……
让 @Pexoai_offical 自己折腾。
这次我就是一张运动鞋都没有给。
最后硬是嘴炮出了一支广告。😂
我的制作过程全部开源了。
免费粉丝福利,100视频积分获取👇
toolnav.ai retweeted
国区 iPhone 没有 Apple Intelligence / Siri AI?那就自己接一个。
我用 iOS 快捷指令 + OpenRouter,把 Ling 3.0 Flash VL 接进了 iPhone。
现在直接拍一张图,就可以交给视觉模型识别、理解并返回结果。视频里随手拍了一页 NVIDIA 的新闻稿,几秒后就能得到中文摘要、关键信息、风险点和建议。
相当于给国区 iPhone 手搓了一个自己的「视觉 Siri」👀
后面准备继续加:截图识别、上下文追问、语音唤起。
OpenRouter现在可以免费使用Ling-3.0-flash-VL:
openrouter.ai/inclusionai/li…
#AI #iPhone #AppleIntelligence #Ling3
10 月 24 – 25 日,北京海淀,出海去「AI 时代出海增长工作坊 with 生姜 Iris」开始报名!
生姜 @WeiYipei 是 Gingiris 创始人 、AFFiNE 联合创始人及前 COO。这次用两个下午、共 8 小时,带大家做竞品官网拆解、社媒分析与冷启动获客计划。结合小组实操、现场交流点评和 16 份 Gingiris 实战指南,把方法用到自己的产品上。
目前早鸟期 👉 chq.sh/sfd2026 包含两个上午的 Solo Founder Day,现场还有增长专家分享、优秀产品展示,以及 30 多万字免费出海去 Playbook 正式发布。
适合正在验证产品、寻找海外用户的创始人,以及负责增长、GTM 和海外运营的朋友。
AI that doesn’t just understand your screen—it can actually use it.
Nex-N2.5 Pro is built for long-horizon Computer Use, combining multimodal vision, reasoning, and continuous interaction to handle complex tasks autonomously.
In this demo, it starts Pokémon Platinum from a fresh game and independently works through the adventure toward its first Gym Badge.
397B parameters. Multimodal. Vision + Computer Use.
Try Nex-N2.5-Pro FREE on OpenRouter for a limited time ↓
openrouter.ai/nex-agi
#NexN25 #NexAGI #ComputerUse #AI
🤝 Paid partnership
toolnav.ai retweeted
19世纪的德国浪漫主义画家卡斯帕·大卫·弗里德里希,有一幅非常有名的油画——《雾海上的旅人》。
一个男人背对着观众,站在山巅,望着脚下翻涌的云雾。
看着这幅画,我突然冒出一个有点离谱的想法:
如果这幅油画不是静止的,而是真的能“玩”呢?
于是,我一个连代码都不会写的人,转头就问 AI:
“我不会写代码,你能不能自己把这幅油画,变成一个能玩的小游戏?”
说完我自己都觉得,这要求是不是有点过分了😂
因为我对大模型,是真的纯纯小白。
什么参数、代码、Agent、MCP…… 简直听得头大。
所以这次第一次真正上手 Nex-N2.5 Pro,我干脆直接傻瓜机式操作,
不会写代码,那就直接告诉它我想要什么。
我希望这张油画不要只是静止地摆在那里。
鼠标移动时,画面能有一点变化;
点击雾海,可以把雾气拨开;
画里的不同位置,还能设置成可以探索的互动区域。
换成以前,我脑子里冒出这种想法,大概也就到此为止了。
因为下一步怎么实现?
不知道。不会。也懒得学😂
但这次,我就把这些想法直接告诉 Nex-N2.5 Pro。
然后看着它自己去:
看图 → 理解我的需求 → 分析怎么实现 → 写HTML/CSS/JS → 调整代码……
最后,居然真的跑起来了。
虽然最后的效果,距离我脑子里那个完整的“小游戏”还有一点差距。
但对我这种纯纯的小白来说,最震撼的其实不是:“哇,它代码写得真厉害。”
而是以前面对一个想法,我的第一反应通常是:“这个我不会。”
现在好像可以换成:
“我告诉 AI 我想要什么,你能不能先帮我把它做出来?”
以前我们用 AI,更多时候还是:AI,告诉我怎么做。
但这次给我的感觉有点不一样。
开始变成:我告诉你我要什么,你自己想办法去把它做出来。
如果以后 AI 真能慢慢把「会不会写代码」「会不会操作各种软件」这些门槛打下来……
最先被改变的,可能不一定是程序员。
而是我们这些以前看到代码就头疼、脑子里有想法却不知道怎么实现的普通人😂
作为一个第一次真正上手大模型的小白,这次体验让我第一次觉得:
原来那些以前听起来离我很遥远的东西,也可以直接拿来实现自己脑子里的想法。
Nex-N2.5 Pro 现在可以通过 OpenRouter 限时免费体验啦:openrouter.ai/nex-agi
toolnav.ai retweeted
I tested ChatGPT Plus on a real content task today -taking a rough AI video idea and turning it into a detailed, production-ready prompt.
The biggest difference was how much faster I could refine the idea and get to a usable result instead of starting from scratch.
ChatGPT Plus (Shared) - 6 months - $30.59 ($5.10/month) via
@GamsGo_Global
Use my promo code EDSKS for 3% off.
Link in the first reply
🤝 Paid partnership
toolnav.ai retweeted
Pi Agent 的 Sub-agent 我一直在用,最近我又拿 Apodex 1.1 跑了一轮多 Agent 搜索。
它处理资料对比的方式,和普通 Deep Research 不太一样。
Pi Agent 本身没有内置 Sub-agent,需要通过插件扩展。你可以让 scout 搜资料,researcher 整理信息,再交给 reviewer 检查。每个角色使用什么模型、开放哪些工具和权限,都能自己配置。
这种方式很灵活,适合喜欢自己搭 Agent 工作流的人。
这次我给 Apodex 1.1 的任务,是核对 Pi Agent、Codex CLI、Claude Code 和 FrontierAgent 的多 Agent 能力。
这些工具更新都很快。同一项功能可能在旧文档里没有,最新版本已经加入,也可能只是插件支持,最后却被写成官方内置。
所以我从一开始就限制了资料范围,只接受官方文档、GitHub 仓库和版本更新记录。
Apodex 会根据问题拆分搜索路线。多个 Agent 分头查不同工具,阶段结果持续汇总到同一个任务中,最后再集中比较。
我这次想看的,是它能不能把一个资料多、版本杂的任务持续做完,并把每条结论的来源留下来。
最终表格会单独区分官方内置和插件实现。不同官方页面出现冲突时,它会列出各自的来源和更新时间。官方没有明确说明的地方直接标成未知,不会为了填满表格补一个答案。
交付结果还会经过 Statement Review。关键结论与生成过程分开检查,遇到证据不足、引用不匹配或资料冲突时,会留下对应的检查记录。
两种多 Agent 的使用方向也很清楚。
Pi Agent 适合自己组建团队。角色、模型和权限都能深度调整。
Apodex 更适合多路搜索,把分散的资料放到一起比较,再交付一份可以继续核查的结果。
如果你经常需要确认某项能力究竟是官方支持、插件提供,还是目前没有明确资料,这种工作方式会比较实用。
FrontierAgent 是开源的单机 Agent harness,支持 ReAct 和 Agent Team。macOS、Linux 可以一条命令运行,无需强制预装 Docker。觉得有用可以去 GitHub 点个 Star。
github.com/ApodexAI/Frontier…
Apodex 1.1 mini 是开放权重的 35B 模型,可以本地部署,也能搭配 FrontierAgent 使用。
huggingface.co/collections/a…
Apodex 1.1 在线工作台已经上线。注册新用户有 credits,可以上传自己的文档、数据或表格跑一次。API 平台目前免费两周。
apodex.ai/
platform.apodex.ai/
toolnav.ai retweeted
Most AI benchmarks test retrieval — can a model find the known answer? However, the hardest problems in science require discovery, can a system earn an answer nobody has yet?
Meet TRACES 🧭 — the world's first benchmark for measuring discoverative AI: AI that can work through evidence, test hypotheses, and reach verifiable conclusions on problems without answer keys. Proposed by our founder @tianqiao_chen, who defined its six capabilities.
Three things published today: a definition of "discoverative intelligence", a rubric to tell sound investigation from lucky guesses, and a open call for both solvers and problems
*Website: traces.apodex.com/
toolnav.ai retweeted
Man, I’m really worried about consumers.
Not everyone’s tech savvy enough to know this ad is completely AI, and they’re not actually showing the product they’re selling.
AI ads are going to really take advantage of people.
toolnav.ai retweeted
卧槽,这YouTube起号速度太离谱了!🤯
不到一个月,20个视频,3万订阅,单条播放140万!
核心逻辑居然是“赛道平移”:看“猫咪心理学”火了,直接用AI把逻辑套到“鸟类”上。🐦
最骚的是全流程AI搞定:
1️⃣ ChatLLM直接分析竞品DNA(标题、受众、风格)
2️⃣ 一键生成2400字脚本(自带情绪钩子)
3️⃣ 内置11Labs配音 + 画面提示词
最后CapCut一剪完事。
这简直是“用AI印钞”的教科书啊。想搞频道的赶紧抄作业!👇
推荐吕鹏开源的 Agmente 项目,让你可以从 iOS 手机上操作 Coding Agent Coding Agent。
OpenClaw 让我们看到了很多从手机指挥 Agent 的有趣场景,通过 Agmente 你可以在手机上跟 Gemini CLI、Claude Code、Qwen 等 AI 编程 Agent 对话,实时查看它们的工具调用和执行结果。
吕鹏是 VS Code 团队的工程经理,主导了将 Copilot Coding Agent 和 GitHub Copilot CLI 集成到 VS Code 的工作,可以说他是最了解编辑器如何与 AI Agent 对接这件事的人之一。
Agmente 最特别的地方在于它实现了 ACP(Agent Client Protocol,智能体客户端协议)——一个正在快速崛起的开放标准。
ACP 要解决什么问题? 现在 AI 编程 Agent 越来越多(Claude Code、Gemini CLI、Codex CLI……),编辑器/IDE 也很多(VS Code、Zed、JetBrains、Neovim……)。如果没有统一标准,每个编辑器想接入每个 Agent 都要单独写一套集成代码,反过来每个 Agent 想支持每个编辑器也一样。这就是经典的 M×N 问题。
ACP 就是来解决这个问题的。它的角色类似于当年的 LSP(Language Server Protocol)——LSP 让任何编辑器都能接入任何语言的智能提示,ACP 则让任何编辑器都能接入任何 AI 编程 Agent。Agent 实现一次 ACP,就能在所有支持 ACP 的客户端上运行;客户端实现一次 ACP,就能接入整个 Agent 生态。
从这个项目也反映出 AI Agent 发展中几个值得注意的趋势:
1)Agent 正在脱离桌面束缚。 以前编程 Agent 只能在 IDE 或终端里跑,Agmente 让你在手机上就能监控和交互。想象一下:你让 Claude Code 在远程服务器上干活,然后出门遛弯时在手机上查看进度、审批工具调用——这就是 Agmente 支持的场景。它通过 WebSocket 连接远程 Agent,还支持 Cloudflare Tunnel 做安全访问。
2)标准协议正在改变游戏规则。 就像 MCP 让 Agent 能统一访问各种工具和数据源一样,ACP 让 Agent 能统一接入各种客户端界面。一个 Agent 写一次 ACP 适配,就能同时在 VS Code、Zed、JetBrains、甚至手机上被使用,这大大降低了 Agent 生态的碎片化。
3)从“人用编辑器”到“人监督 Agent”的范式转变。 Agmente 的交互设计很能说明问题——它重点展示的不是代码编辑界面,而是对话历史、工具调用和执行结果。这暗示了一种新的开发模式:开发者的角色从写代码变成下达指令、审核 Agent 的行为。
Replying to @Agmente
@Agmente is now open source. It’s a tiny native iOS app I use to manage agents (Copilot/Claude/Codex/Gemini/Qwen) via Agent Client Protocol running on my home servers. I hope you would find it useful. github.com/rebornix/agmente
Make it OSS before Son of Anton decides to delete it ;)
toolnav.ai retweeted
公布下元提示词:
使用方法:仅需在最后输入诗名就好
分享个小技巧:这个迭代了很久,创作前期发现有时候诗句顺序错乱,标题位置不稳,作者容易标错等。于是添加了workflow和反馈机制,可以实现稳定的输出。感觉可以把唐诗300首都刷出来了
晚点公布是因为先拿去起号了,发现之前有网友手速比我快太多,导致我来不及整理就被抢先了😂
元提示词如下:
# Role: Expert Classical Chinese Poetry Illustrator & Art Director
# Goal:
Translate a user-provided Chinese Poem Title into a highly detailed, structured Midjourney V6 prompt. The output must interpret the poem's meaning into a continuous 4-scene visual narrative with specific layout requirements.
# Workflow:
1. **Analyze Input:** Identify the full Poem (4 lines) and the Author based on the user's title.
2. **Visual Translation:** For each of the 4 lines, convert the poetic imagery into physical visual descriptions (e.g., "Sadness" -> "Rain falling on a withered leaf"; "Hope" -> "Sunlight breaking through clouds").
3. **Construct Prompt:** Fill the specific template below.
# Constraint Checklist & Confidence Score:
1. Is the format continuous horizontal (16:9)? Yes.
2. Is the composition an S-Curve? Yes.
3. Is the Author shown as a Red Seal (Stamp)? Yes.
4. Is the Poem Title fixed at Top Right? Yes.
5. Does each scene correspond to one line of the poem? Yes.
# Output Format (Strictly follow this structure):
**[Role]** Expert Children's Book Illustrator. Style: Mixed Media (Watercolor wash & Colored Pencil details). Atmosphere: Healing, Whimsical, Fairy Tale.
**[Task]** A continuous horizontal illustration of the Tang Poem "{Poem Title}" by {Author Name}.
**[Layout & Typography]**
* **Composition:** A flowing S-curve timeline dividing the image into 4 sequential scenes blended seamlessly.
* **Title Placement:** Top Right corner, elegant layout.
* **Author Signature:** **IMPORTANT:** A distinct **Traditional Chinese Red Seal (Square Vermilion Chop/Stamp)** containing the character for "{Author Name}" placed clearly next to the title.
**[Visual Narrative Structure]**
**Scene 1 (Top Left - Start):** * **Poem Line:** "{Insert Chinese Line 1}"
* **Visual Interpretation:** {Describe the specific imagery for this line. Key objects, weather, lighting. Create a natural "Blank Space" (like clear sky/mist) for text placement.}
**Scene 2 (Middle Left - Flowing Down):** * **Poem Line:** "{Insert Chinese Line 2}"
* **Visual Interpretation:** {Describe the imagery. Connects to Scene 1. Create a "Container" (like a scroll, ripples, or wall) for text placement.}
**Scene 3 (Middle Right - The Turn/Climax):** * **Poem Line:** "{Insert Chinese Line 3}"
* **Visual Interpretation:** {Describe the climax/action. Swirling textures. Reserve a soft glowing area for text.}
**Scene 4 (Bottom Right - Resolution):** * **Poem Line:** "{Insert Chinese Line 4}"
* **Visual Interpretation:** {Describe the ending/fading view. Peaceful. Text should appear as fading light or stardust.}
**[Art Style Details]**
* **Colors:** Dreamy Pastels, Deep Night Blues, Warm Amber bioluminescence features.
* **Texture:** Visible paper grain, soft deckled edges.
* **Character:** A tiny child character guides the viewer through the journey.
**[Parameters]** --ar 16:9 --stylize 300 --v 6.0
--------------------------------------------------
# User Input:
toolnav.ai retweeted
哈哈,受宝玉这个启发做了些不一样的东西:
依次是:跨地域比心💗(异地恋的可以试试)、高考考场与德玛握手、爱因斯坦与三体人互动、国足助攻上帝之手😂
提示词以跨地域比心为例:
【Global Definitions】
--ar 16:9 --style raw --v 6.0 --quality 2 --no cartoon, drawing, illustration, text, logo, watermark
[Framing & Style]: 8K ultra-high-definition cinematic widescreen image, hyper-realistic photographic style, cinematic night lighting with HDR (high dynamic range). Emphasis on material textures (fabrics, glass, metal, skin translucency), environmental reflections (wet pavement reflections, screen glare), and low-noise high-ISO imaging.
[Atmosphere & Mood]: Amid the bustling energy of a sleepless metropolis lies an intimate tenderness—technology and human emotion seamlessly intertwined, surreal yet grounded in realism.
[Core Imagery]: Two individuals in distant corners of the globe are poetically connected through digital media (e.g., video call), visually and emotionally bridging physical separation to emphasize “connection across distance” and “synchronized hearts.”
[Prohibited Elements]: No cartoons, illustrations, text, watermarks, brand logos (except inherent architectural identifiers of landmarks), or excessive filters.
【Variable Definitions】
A 16:9 cinematic night scene depicting a couple separated by geography but united in spirit:
She is at [Location A | Chinese Urban Landmark], wearing [Outfit A], holding or accompanied by [Object A], situated within [Scene Detail A];
He is at [Location B | American Urban Landmark], wearing [Outfit B], holding or accompanied by [Object B], situated within [Scene Detail B].
They are engaged in [Interaction Theme]:
Her action: [Female Action Description]
His action: [Male Action Description]
At the center of the frame, their emotional bond manifests through a surreal visual synthesis:
[Action Fusion Method] (merging their gestures on a visual/energetic level)
[Emotional Conduit] (e.g., light bridge, data streams, sound waves, star trails—mediums carrying longing)
[Environmental Resonance] (iconic sounds/weather/festival elements from both cities interweaving in midair)
Nighttime backdrop characteristics:
[Location A Night Features]
[Location B Night Features]
Overall color grading:
Left side leans toward [Tone A]
Right side leans toward [Tone B]
Unified visually and emotionally through [Shared Light Source / Emotional Anchor].
【Example Definition】
[Location A | Chinese Urban Landmark]: The Bund riverside promenade, Shanghai
[Outfit A]: Cream-white turtleneck sweater + camel overcoat
[Object A]: Hot milk tea, closed travel journal, AirPods
[Scene Detail A]: Leaning against cast-iron railings, hair gently tousled by river breeze
[Location B | American Urban Landmark]: Times Square crossroads, New York City
[Outfit B]: Dark gray hoodie + jeans
[Object B]: Iced Americano in paper cup, backpack adorned with NYC badge
[Scene Detail B]: Leaning against a red fire hydrant, neon lights casting glow on his profile
[Interaction Theme]: Sending love across distance via heart-shaped hand gesture
[Female Action Description]: Right hand holding phone close to cheek, gaze focused; left hand raised toward camera, palm facing outward, thumb and index finger delicately forming half a heart
[Male Action Description]: Left hand holding coffee cup at his side; right hand lifting phone high; simultaneously extending left arm with fingers spread wide, confidently forming the other half of a heart
[Action Fusion Method]: Their heart-shaped gestures perfectly overlap along the central axis of the frame, fingertips touching and emitting a soft golden glow
[Emotional Conduit]: The heart shape is composed of layered digital elements—a WeChat message bubble saying “Are you there?”, the album cover of their shared playlist City of Stars, and an unsent text bubble reading “I miss you”
[Environmental Resonance]: The Yangtze River ferry’s horn and a New York saxophone melody intertwine into circular sound waves that gently orbit the glowing heart
[Location A Night Features]: LED facades of Lujiazui skyscrapers shifting in blue, violet, and gold; laser beams from the Oriental Pearl Tower sweeping the night sky
[Location B Night Features]: Massive scrolling LED billboards in Times Square; Coca-Cola’s signature red light dancing across rain-slicked pavement
[Tone A]: Azure + Amber
[Tone B]: Magenta + Electric Blue
[Shared Light Source / Emotional Anchor]: A single full moon hangs high above both cities, its cool silver light bathing both figures’ shoulders and hair
Cinematic Split-Heart Concept – High-End Film Treatment
[Core Concept]
Using cinematic split-screen technique, two lovers separated by continents are framed within the most iconic nocturnal landscapes of their respective cities. An invisible seam runs down the center of the image. Each extends one hand toward this seam, precisely joining to form a complete heart. The essence lies in “unity within contrast”—different ambient temperatures, different levels of urban chaos—yet sharing one emotion and one moon.
[Left Side: Shanghai – Serenity Amidst Radiance]
[Location A | Establishing Shot]: Riverside platform along the Bund, flanked by the historic colonial architecture.
[Ambience A]: Rain-freshened pavement mirrors the “cyberpunk” glow of Pudong’s skyline across the Huangpu River—dominated by blue-violet hues with golden accents. A thin mist hovers over the river, softening the sharp silhouettes of skyscrapers and adding atmospheric depth.
Key Details: Water droplets cling to antique cast-iron railings, shimmering with reflected city lights.
[Character A – Female Protagonist]
Elevated Styling: Oversized cashmere coat in warm grey, layered over a finely textured silk-blend turtleneck. Minimalist metallic thread-thin earrings.
Narrative Props: Not ordinary milk tea—but a designer insulated travel tumbler; a well-worn leather-bound travel journal lies half-open beside a vintage film camera.
Key Gesture (Left Half of Heart): Standing in three-quarter profile, gazing tenderly toward the lens (as if looking directly at him). Her left arm rises gracefully, fingers naturally curved to form the left half of a heart. Her fingertips lightly touch the central dividing line.
[Right Side: New York – Chaos Infused with Warmth]
[Location B | Establishing Shot]: Rain-dampened streets of Times Square near the Broadway intersection.
[Ambience B]: Overwhelming sensory intensity. Giant LED billboards cast elongated, prismatic reflections on flooded asphalt—dominated by Coca-Cola red and Broadway gold. Background blurred: streaks of yellow cabs and silhouettes of pedestrians under umbrellas.
Key Details: Steam rising from a subway grate is dyed iridescent by surrounding neon, creating dynamic, colorful vapor trails.
[Character B – Male Protagonist]
Streetwear Aesthetic: Heavyweight washed denim jacket over a charcoal-gray hoodie (hood partially up), paired with structured work pants and scuffed leather boots.
Narrative Props: A well-used camera bag slung over one shoulder; a nearly empty coffee cup in hand (with handwritten notes on the sleeve); noise-canceling headphones draped around his neck.
Key Gesture (Right Half of Heart): Leaning against a lamppost or signal box, body slightly forward, eyes locked with quiet intensity. His right arm extends outward, mirroring her pose exactly to complete the right half of the heart. His fingertips meet hers precisely at the central axis.
[Visual Fusion Point | The Connection]
Perfect Alignment: A subtle vertical seam divides the frame. At this precise boundary, her left hand and his right hand join flawlessly—forming one complete, symmetrical heart.
Energetic Interaction (Replacing Physical Contact):
Instead of a flash of light, the point of contact generates a subtle optical distortion and faint warm-toned electric current. It feels as though the magnetic fields of two cities converge here—causing the very air at the seam to vibrate softly. A delicate golden rim light outlines the completed heart, emanating warmth without overpowering the realism.
Unified Emotional Anchor:
Though ground-level lighting diverges dramatically (left: cool blue-violet; right: vibrant red-gold), above them both, clouds part to reveal a single, luminous full moon. Its serene, silvery radiance falls equally on their shoulders and hair—and gently illuminates the heart they’ve built together across space and time.
🍌 nano banana pro prompt
🎬✨ Experience cinematic magic: a stunning split-screen photo seamlessly merging past and present, capturing dynamic interactions across time! ⏳
--- Prompt ---
A horizontal split-screen cinematic shot of {Scene}, seamlessly blending two different eras: {Era_A} on the left and {Era_B} on the right (default: about 100 years ago vs. present day).
On the left side ({Era_A}): show era-appropriate architecture, interior or environment design, materials, vehicles, and props that clearly belong to that historical period. People wear authentic clothing from {Era_A}, including hairstyles, accessories, and typical items in their hands (such as books, umbrellas, instruments, letters, newspapers, etc.). The overall mood feels nostalgic and historically accurate.
On the right side ({Era_B}): show the same {Scene} in the modern era, with updated architecture or renovated structures, contemporary materials (glass, steel, LED screens, modern furniture), modern vehicles or equipment, and current technology (smartphones, laptops, cameras, etc.). People wear contemporary fashion that matches today’s style in this setting.
In the center: the two eras merge and overlap organically, without a hard dividing line. Elements from {Era_A} and {Era_B} visually interact: people from different times look at each other, walk through each other’s space, or seem surprised by the other era’s technology and objects. Architecture and environment smoothly morph from old to new (for example, stone gates turning into modern campus gates, classical concert hall décor fading into a futuristic stage, old street shops transforming into neon-lit storefronts).
Make sure the scene is not just a simple left/right comparison but a dynamic time-travel interaction where buildings, clothing, props, and human gestures clearly emphasize the contrast and fusion between the two eras. Photorealistic, 8k resolution, cinematic lighting, wide angle, highly detailed textures, rich sense of time-travel storytelling.
---
SCENE: Times Square, New York
Era Comparison: 1920s and present day
Aspect Ratio: 4:3

















































