Did they still not release "Bring your own subscription" "login with ChatGPT" and letting people apply their subscription/quota to other apps/services? There are so many use-cases where someone builds a usefull scaffold (e.g. even static website) that could benefit from integration with LLMs but rolling the authentication, having API-keys, free budgets etc. is scary. It would be much better if that could be outsourced to the LLM provider of the user.
It would bear a good lock-in effect for OAI as well
Neither Anthropic or OpenAI want to enable their fully self-service, automated, and scalable enterprise feature flags for only, say, 50 seats each burning $3k a month.
It's particularly annoying to see these feature flags exist in the UI; the features are automated, yet they won't let you select a plan above their "Team" level plans to use them.
(Over on the API side, they're more usage based, and easy to hit the highest tier from zero within about 10 weeks, but certain "talk with support because we probably have to edit a config file we don't let the riffraff see there's a setting for" features, like flipping off a safety flag so the LLM won't refuse to analyze a contract, are similarly gated. That seems more reasonable than gating infosec.)
On the end-user app side, table stakes data security features, some of which used to be included for lower level accounts then just got toggled off, are gated behind somewhere on the order of 150 seats at greater than Max usage instead of 25 or 50 seats.
Would be great if they provided at least some guidance how to keep this thing on topic. Even the official demo https://chatkit.world/ is not restricted, it happily chats about whatever.
OpenAI wants all these services to pay per request. That is their entire business model. There is no incentive for them to fold everything into the same $20/mo subscription (which they are most definitely selling at a loss).
It might meaningfully change the business model of LLM businesses. Becomes seemingly much harder to universally charge $30/mo subscriptions when the user is bringing their own API key.
IMO it is usually a mistake to integrate AI as a separate chat function in the UI. AI(s) be integrated into existing chat and collaboration features via @ tagging. E.g. in something like Figma or Google Docs, you should be able to add comments, tag @ai, and ask it for feedback or to make changes just as you would if you were collaborating with a human. Putting the AI inside its own private chat button often makes it feel like a Clippy-esque gimmick.
Solid insight - I actually tried implementing an AI chat into my app. It's just an internal app for my own productivity purposes with my eCommerce business, but I felt I had a credible use case for having it there. At the end of the day, the work required to get it to a spot where it was useful and not janky was just too much (even with the AI tooling provided by Vercel's AI SDK). And, even though I got it working, I never found a reason to use it that I didn't feel like could just be achieved more smoothly with some buttons and text inputs.
Interesting how they claim to be "framework-agnostic" but only support React and don't really integrate with any backend framework. I've been building something that helps create proper framework-agnostic libraries[0], I even use AI chat as an example.
It supports vanilla Js, so that covers all frameworks. I see your pain, I built predictabledialogs.com and didn't expect OpenAI to do a no code builder and a chat widget. eeks I was hoping they will aim for AGI while we do the boring work.
Maybe I'm being dumb but is this a generic chat UI for openai models only? Pretty bearish on adoption of this if so. As a pragmatic dev I'd definitely not be keen to bake model lock-in into my UI for functionality as generic as chat.
Is there a place to change the endpoint url? It seems we just add the workflow id and generates a secret which is used by the frontend. Apache license is good though
Note that their basic example in the readme starts with `OpenAI(api_key=os.environ["OPENAI_API_KEY"])`
Generally speaking, in most regular usages, you can replace it with an alternative provider like Openrouter, with `OpenAI( base_url="https://openrouter.ai/api/v1", api_key="<OPENROUTER_API_KEY>")`
But I haven't tested chatkit yet and don't know if it might be using special endpoints that are currently only supported by OpenAI. IANAL, but would assume though that if the client is Apache licensed, then it wouldn't be an issue for Openrouter and others AI providers to develop their own versions of those endpoints.
Maybe I'm overly optimistic, but based on what I'm seeing with MCP and other recent developments, the industry is continuously gravitating towards a commoditized/interchangeable future where no provider has a structural moat.
But even with AG-UI,CopilotKit is a paid service. Everyone's trying to do vendor lock in, it's really annoying. They make the dev experience worse and you have to pay for it? It's like every langcrap service out there
And CopilotKit needs a JS backend for a proxy that holds the state for the chat which is quite a pain if you want to scalably self-host. Maybe ChatKit will be less lock-in than CopilotKit.
I'm working on https://github.com/theblazehen/react-ag-ui which implements the ag-ui protocol, however your solution looks significantly more polished. I'd be interested in potential collaboration,
I hope you were not one of those who applied for the startup accel at openai, it seems they just pulled the rug from under those startups by building what they were.
There's barely any code in the js bindings anyway. It's just a thin component which defines an interface to communicate with it via events. The logic all happens in the (AFAIK) closed source iframe that gets injected from https://cdn.platform.openai.com/deployments/
Smart for OpenAI to launch, from a product lock-in perspective, since most similar chat sidebars in apps tend to let one choose the AI model from a list that always seems to include Claude as well....
But for that same reason, not sure I'd want to implement this into products since it does lock you into OpenAI, only.
There are so many libraries called chatkit or some variation of that including the one I am working on - chatbotkit (I guess some SEO will rub off).
Apart from that, it looks like an interesting initiative, but I fail to see what this is supposed to replace. Widgets? Customer support? I mean, sure, you can create some kind of question-answer bot, but for something more substantial, you'll need a completely different dashboard and backend. I can't see how OpenAI will accomplish this without butchering their existing UI.
their chatkit demo (chatkit.world) doesn't even work in safari. also "react bindings" and "vanilla js" isn't exactly "framework-agnostic". I don't like the idea of embedding someone else's UI in my app. This seems like an overly generalized solution for a specific set of problems.
Did they still not release "Bring your own subscription" "login with ChatGPT" and letting people apply their subscription/quota to other apps/services? There are so many use-cases where someone builds a usefull scaffold (e.g. even static website) that could benefit from integration with LLMs but rolling the authentication, having API-keys, free budgets etc. is scary. It would be much better if that could be outsourced to the LLM provider of the user.
It would bear a good lock-in effect for OAI as well
Thats like the opposite of what Open AI would really want though.
They want people walking up saying "we commit to spending $100k monthly for the next year let's discuss bulk discounts."
Neither Anthropic or OpenAI want to enable their fully self-service, automated, and scalable enterprise feature flags for only, say, 50 seats each burning $3k a month.
It's particularly annoying to see these feature flags exist in the UI; the features are automated, yet they won't let you select a plan above their "Team" level plans to use them.
(Over on the API side, they're more usage based, and easy to hit the highest tier from zero within about 10 weeks, but certain "talk with support because we probably have to edit a config file we don't let the riffraff see there's a setting for" features, like flipping off a safety flag so the LLM won't refuse to analyze a contract, are similarly gated. That seems more reasonable than gating infosec.)
On the end-user app side, table stakes data security features, some of which used to be included for lower level accounts then just got toggled off, are gated behind somewhere on the order of 150 seats at greater than Max usage instead of 25 or 50 seats.
Would be great if they provided at least some guidance how to keep this thing on topic. Even the official demo https://chatkit.world/ is not restricted, it happily chats about whatever.
Oh man this thing tells me my iPhone 13 mini isn’t wide enough to use it, that’s rough.
It doesn’t work on my iPhone 17 Pro Max either and when I turn it to landscape it’s buggy and doesn’t work. Great demo…
Doesn't work either on my Samsung s25
OpenAI wants all these services to pay per request. That is their entire business model. There is no incentive for them to fold everything into the same $20/mo subscription (which they are most definitely selling at a loss).
It might meaningfully change the business model of LLM businesses. Becomes seemingly much harder to universally charge $30/mo subscriptions when the user is bringing their own API key.
IMO it is usually a mistake to integrate AI as a separate chat function in the UI. AI(s) be integrated into existing chat and collaboration features via @ tagging. E.g. in something like Figma or Google Docs, you should be able to add comments, tag @ai, and ask it for feedback or to make changes just as you would if you were collaborating with a human. Putting the AI inside its own private chat button often makes it feel like a Clippy-esque gimmick.
Solid insight - I actually tried implementing an AI chat into my app. It's just an internal app for my own productivity purposes with my eCommerce business, but I felt I had a credible use case for having it there. At the end of the day, the work required to get it to a spot where it was useful and not janky was just too much (even with the AI tooling provided by Vercel's AI SDK). And, even though I got it working, I never found a reason to use it that I didn't feel like could just be achieved more smoothly with some buttons and text inputs.
haha, shameless plug but if you use Google Sheets you can try adding `try@tabtabtab.ai` to the sheet as an editor, tagging it and giving it work
Great ! The strategy to make your complementary product easy to create
https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/Interesting how they claim to be "framework-agnostic" but only support React and don't really integrate with any backend framework. I've been building something that helps create proper framework-agnostic libraries[0], I even use AI chat as an example.
[0] https://fragno.dev/
They don't only support React. There are examples in their guide for how to do all theming with vanilla JS. And you can use any backend you want.
It supports vanilla Js, so that covers all frameworks. I see your pain, I built predictabledialogs.com and didn't expect OpenAI to do a no code builder and a chat widget. eeks I was hoping they will aim for AGI while we do the boring work.
I think fragno proposition is sleek.
Maybe you could use your software to help them support other frameworks
> framework-agnostic, drop-in chat solution
Maybe I'm being dumb but is this a generic chat UI for openai models only? Pretty bearish on adoption of this if so. As a pragmatic dev I'd definitely not be keen to bake model lock-in into my UI for functionality as generic as chat.
The openai library works with most other providers by just changing the endpoint url, and this is Apache licensed, so I feel good about using it.
Is there a place to change the endpoint url? It seems we just add the workflow id and generates a secret which is used by the frontend. Apache license is good though
Note that their basic example in the readme starts with `OpenAI(api_key=os.environ["OPENAI_API_KEY"])`
Generally speaking, in most regular usages, you can replace it with an alternative provider like Openrouter, with `OpenAI( base_url="https://openrouter.ai/api/v1", api_key="<OPENROUTER_API_KEY>")`
But I haven't tested chatkit yet and don't know if it might be using special endpoints that are currently only supported by OpenAI. IANAL, but would assume though that if the client is Apache licensed, then it wouldn't be an issue for Openrouter and others AI providers to develop their own versions of those endpoints.
Maybe I'm overly optimistic, but based on what I'm seeing with MCP and other recent developments, the industry is continuously gravitating towards a commoditized/interchangeable future where no provider has a structural moat.
Looks like it supports custom backends
This looks very similar to CopilotKit and AG-UI. With CopilotKit you can have some tools implemented client-side and others server-side using AG-UI.
https://www.copilotkit.ai/
With ChatKit you can proxy everything through your ChatKit API "proxy" and implement the "agentic" stuff in the proxy.
Even the React APIs look very similar.
Looks like a vendor lock-in and a way to stop the AG-UI ecosystem
But even with AG-UI,CopilotKit is a paid service. Everyone's trying to do vendor lock in, it's really annoying. They make the dev experience worse and you have to pay for it? It's like every langcrap service out there
And CopilotKit needs a JS backend for a proxy that holds the state for the chat which is quite a pain if you want to scalably self-host. Maybe ChatKit will be less lock-in than CopilotKit.
I work on assistant-ui.com
MIT and no proprietary cloud/proxy service :)
We do sell a chat history service if you don’t want to manage your own database though
Have you looked at using AG-UI?
I'm working on https://github.com/theblazehen/react-ag-ui which implements the ag-ui protocol, however your solution looks significantly more polished. I'd be interested in potential collaboration,
I hope you were not one of those who applied for the startup accel at openai, it seems they just pulled the rug from under those startups by building what they were.
deep-chat is a solid chat UI
https://github.com/OvidijusParsiunas/deep-chat
> framework-agnostic
Later,
> 2. Install the React bindings
So not so framework-agnostic, or am I missing something ?
There's barely any code in the js bindings anyway. It's just a thin component which defines an interface to communicate with it via events. The logic all happens in the (AFAIK) closed source iframe that gets injected from https://cdn.platform.openai.com/deployments/
They have web components if you want
[dead]
no demo link or screenshots in the Github readme? OpenAI surely could have had their marketing department create something quick?
Demos are at https://chatkit.studio/
> Your screen is too small for chatkit.world
> Make your window wider or open on desktop to continue
For a chat app UI?
(BTW, I had to take a screenshot then use image to text inference in order to copy that message)
Here it is in all it's glory (Pixel 7 Pro): https://imgbox.com/EUhc4iVi
Smart for OpenAI to launch, from a product lock-in perspective, since most similar chat sidebars in apps tend to let one choose the AI model from a list that always seems to include Claude as well....
But for that same reason, not sure I'd want to implement this into products since it does lock you into OpenAI, only.
They badly need more sticky adoption, openai specific workflows are the play for them (docgpt, this automation no-code, etc).
There are so many libraries called chatkit or some variation of that including the one I am working on - chatbotkit (I guess some SEO will rub off).
Apart from that, it looks like an interesting initiative, but I fail to see what this is supposed to replace. Widgets? Customer support? I mean, sure, you can create some kind of question-answer bot, but for something more substantial, you'll need a completely different dashboard and backend. I can't see how OpenAI will accomplish this without butchering their existing UI.
What are the benefits of this framework compared with alternatives, and will it work with frameworks other than React?
I wonder how it compares to https://ai-sdk.dev
Looks a little nicer than the vercel equivalent.
vercel is better, you can choose any provider you want. not locked in to a single provider. Apache license
their chatkit demo (chatkit.world) doesn't even work in safari. also "react bindings" and "vanilla js" isn't exactly "framework-agnostic". I don't like the idea of embedding someone else's UI in my app. This seems like an overly generalized solution for a specific set of problems.
it doesn't work in chrome either
lost me at `npm install …`