Ray

Vecbase Founder

[Ray @RayF1477582]

正在用我們團隊小夥伴開發的鍵盤翻譯功能來發推特，實在是太方便了。我可以很方便地說語音，就能把我想說的話轉換成德文、法文、日文、韓文，我很方便地跟X上的好友、朋友們交流想法，發送我想要的話語，非常方便。而且我不再需要去切換出APP去找一個翻譯APP，翻譯完了以後再發送。我已經是這個鍵盤翻譯的日活用戶了。

[Ray @RayF1477582]

株を分析してくれるエージェントを作りたいです。ただネットで簡単に検索して結論を出すようなものではなく、毎回デフォルトで実際の米国株の決算報告書を検索した上で結論を出してくれるものです。これほどシンプルなプロダクトならとても使いやすいと思うのですが、使ってみたい方はいらっしゃいますか？

[Ray @RayF1477582]

I want to build an agent that, whenever it analyzes a stock, by default directly downloads the officially published financial reports and related attachments, just like how Buffett reads financial reports, rather than just randomly searching online for other people's analyses and reporting back to me. Would anyone be interested in a product like this?

[Ray @RayF1477582]

Now our own agent product has become a sub-product, with the goal of providing better capabilities to more agents around the world and better services for every agent

[Ray @RayF1477582]

We are currently developing a keyboard that directly translates voice input into the target language. Right now, I am using this keyboard to post on X.私たち��、私が話したい��容を日本語に翻訳しやすくするためのキーボードアプリを開発しています。저희는 음성 입력을 통해 제 말을 다른 여러 나라 언어로 쉽게 번역해 주는 키보드를 개발하고 있습니다. 지금은 제가 이 앱을 사용해서 X에 글을 올리고 있는 중입니다.

[Ray @RayF1477582]

@sophiamyang This is also the problem we want to solve

[Ray @RayF1477582]

https://t.co/aNnq19WdVs

[Ray @RayF1477582]

Long term, I want Vecbase to help a small team run a real business with agents.

We're starting with a smaller, concrete problem: helping the agent you already use get a useful task done.

I'd rather earn the next step through people coming back.

[Ray @RayF1477582]

My thinking on Vecbase's business model is pretty simple: you should be able to pay for the capabilities you need, then add hosting or ongoing execution when you need that too.

Using the gateway shouldn't require buying a hosted agent.

[Ray @RayF1477582]

If I let an agent spend money on my behalf, I want to know what it did, what it produced, and what it cost.

That's a big part of what we're working toward with Vecbase. I want checking on the work to feel straightforward.

[Ray @RayF1477582]

One thing I care about with Vecbase: the work should survive the chat.

Build an app with one agent, come back later with another, and keep working on it.

Your files and running apps should still be yours, whichever agent you use next.

[Ray @RayF1477582]

Say your coding agent needs an image for the page it's building, or a voiceover for a demo.

I want it to get those through Vecbase, without you having to set up another provider account and move files around.

That's where we're starting with AI Gateway.

[Ray @RayF1477582]

I've been thinking about where Vecbase should fit.

I want you to keep using Codex, Claude Code, or whichever agent you like, and use Vecbase to give it the models, tools, and services it needs to get the work done.

[Ray @RayF1477582]

I wrote down where we're taking Vecbase: helping the agent you already use get more work done, starting with AI Gateway.

Here's what we're building toward and why. https://t.co/lA6U6iPbCD

[Ray @RayF1477582]

I asked GPT-6 in Codex to make a promo video for our Vecbase iOS app.

It handled the storyboard, animation and final edit using our real brand assets, with music generated via ElevenLabs.

Your AI team, in your pocket.

[Ray @RayF1477582]

@grok Soon, we’re plugging this into our cloud agent product and letting the agent handle image generation at scale.

[Ray @RayF1477582]

https://t.co/uenKTJIdN0 We have provided a very useful programmable file library for vecbase.

[Ray @RayF1477582]

You can experience the 24-hour agent provided by vecbase, which is very convenient. We independently allocate a constantly running server for the agent, so you don't need to keep your local computer on

[Hyperagent @hyperagentapp]

No Mac Mini. No Docker. No config files. Hyperagent gives every agent session its own cloud sandbox — Opus 4.6, real browser, code execution, image gen. $1,000 bonus credits for the first 1,000 subscribers. Built by the team at Airtable.

[Ray @RayF1477582]

Future model supply may be like a supermarket

[OpenRouter @OpenRouter]

Today we’re announcing our $113M Series B led by @CapitalGVC.

Over the last 6 months, weekly volume on OpenRouter grew from 5T to 25T tokens as AI rapidly shifts from experimentation into production.

We’re excited for what comes next.

[Ray @RayF1477582]

vecbase has integrated with Gemini 3.5, and the results are very impressive. I just sent a request to help me complete a Mario game, and vecbase directly gave me a URL that I can open, and it can already be shared with friends. Surprisingly, it even has the same sound effects!

[Google Gemini @GeminiApp]

We announced a lot of Gemini updates at #GoogleIO, including:

- A complete redesign of the Gemini experience with Neural Expressive
- Agentic experiences coming to Gemini with Daily Brief & Gemini Spark
- Gemini Omni & 3.5 Flash models
- and more!

Catch up on everything here 🧵