Wayfinder Router: deterministic routing of queries between local and hosted LLM
github.comWe need LLM query routing at the OS level like Mobile data. I know it will sound crazy but hear me out. I think about this AI inference as infrastructure. I do not want to pay for it on every app I use it on. I do not think "I have to pay the mobile data of youtube, and the mobile data of whatsapp etc.". I pay Mobile data infrastructure and let my device route it appropiately. In fact, if we ever go the local llm route, you could have LLM capabilities without having access to the internet (or local LAN), and your OS/computer is the only one capable of doing that routing for you.
It doesn't sound crazy at all, this seems almost obvious. The OS should provide a chat completions server and the user should be able to select the underlying LLM's server. This should be just like selecting a default search engine or browser.
Hopefully the EU forces US tech giants to do this. God knows Apple and Google won't do this on their own. They gotta get that sweet default provider revenue.
> It doesn't sound crazy at all, this seems almost obvious. The OS should provide a chat completions server and the user should be able to select the underlying LLM's server. This should be just like selecting a default search engine or browser.
I wonder why this hasn’t happened yet. If Microsoft wants to have a Copilot button and AI investments are all the rage now, surely anything to make integrating with them would be good for keeping the hype cycle alive for longer?
Because it’s still pretty early in this journey and there’s some exclusivity deals to be made.
Honestly I don't get the point but if you want to explore that, both on desktop, mobile or headless server Linux allows you to try it.
You can run ollama with whatever you want on a Debian in literally minutes. You can even do that within a virtual machine using e.g. QEMU, so that you can do all the tests you need risk free.
Again I don't understand what that would enable that can't be done today but it's perfectly fine, you can try today anyway, no need to ask permission to anyone.
No, what I am saying simply does not exist yet.
I am saying I want my OS to expose APIs like it does for the disk or the network for AI. And I want my apps to be able to use those APIs.
I want my backend LLMs to be able to change on a whim. Imagine an Android app consuming from these LLMs. Maybe I am outside and it is making queries to Gemini. And maybe I get home and now it makes queries to my local llm, almost like connecting to local Wifi.
What I am saying does not exist on many levels:
- Agreed upon APIs for this don't think exist (in text maybe, but not in image/sound/video).
- OSs do not expose this (I am not talking manually configured user space stuff here).
- I see a world where your Network provider bundles "calls + data plan + AI tokens". But not only are the offerings for these not standardized, in order to even reach that point we would need to standardize the offerings. How do you compare intelligence among models? How do you compare cost?
- The apps need to start adopting this model
The tech is here, the ecosystem is not.
Well… it doesn’t exist FROM APPLE or MICROSOFT or GOOGLE at their shipped OS Level, but… fundamentally this isn’t a “true OS” level feature you’re asking for, it’s something you think the OS products should bake in, and you might be right! But I think the parents post is suggesting YOU CAN BUILD a prototype of what you want, how it should work, on Linux…
I have a project somewhat close to this I’ve put on pause the last month or so, partly because I’m not sure how useful it is or where to take next, but I may incorporate Wayfinder into it as a next step to improve its capabilities, as part of what it is a model gateway/router that this feels like could make more powerful/flexible in its decision making. I can’t decide if what I’m building is mostly a model recipe cookbook/platform, or a debugging tool, or both or something else at the moment, but, it can do most of that… maybe it’s part of what you want, if you figure that out better? feedback welcome! https://wardwright.dev/ https://github.com/bglusman/wardwright
Why do we need API endpoints? We have the best API there is - the CLI
I mean, the reason mobile data is part of the OS is because the antenna is hardware that must be shared across processes. Chat completions is just a network call like anything else—it’s already available to every app; they don’t need to pay separately (they can use the same account), they just pass their API key over the network to the completions server. What am I missing?
Some kind of routing prompts to different models does make sense. But the usecase of saving money on simple prompt.. I think that has only a slight benefit. Fix my typo doesn't use many tokens anyhow.. also model switching still requires carrying over context so it does have some overhead right.
we could use some composability.
today any kind of routing requires implementing an http proxy to put in the middle
ideally harnesses would support a routing plugin which receives the new whole context and returns just where to send it, and the harness does that. no http proxy. obviously some complications if you want to route from codex to anthropic or openrouter.
but we need to decouple the context building and routing decision from the actual http requests sending, we need to be able to insert "context/routing plugins" in the chain
I'm still waiting for an isolated protocol so we don't have to run the hanress directly on any of the code base's infrastructure. Something as simple as piping everything into and out of an ssh shell would be better than anything I've tested so far.
i've created the protocol, role-model: https://github.com/try-works/role-model
It's funny how much that first paragraph is Claude's voice. I don't know how it got trained so hard to use, "the shape of" for everything.
Loads of ed sheeran in the training data?
Do you want the honest answer?
I'm not sure I understand what this is trying to solve?
If a prompt I give routes to one model, and then another prompt to another model, how does one tie the context together such that the next model knows what's going on?
Otherwise this would only be useful for one-off prompts as far as I can tell.
And if it did keep a context to be passed around, it would always land hot (not in the cache).
Every turn of a conversation with an LLM is getting the whole conversation. Caching complicates the picture, but not by a huge amount. That's why a short question at the end of a long conversation chews tokens faster than it would in a fresh session.
So, a conversation that's ongoing with one model then switching to another would presumably send the whole conversation and the new question. Which defeats the purpose of splitting traffic...so, you're not wrong to question how this actually improves things for anything other than short sessions, which you could choose your own model for if it's a small problem.
sending the whole conversation to a cheap model could still be cheaper than sending just the latest message to the expensive one
you could even take this into account automatically to help decide
Here's a use case: You want to extend the GPT 5.5 quota in you Codex subscription by routing some % of requests to DeepSeek V4 Pro. A router needs to figure out which requests to route where, for the appropriate difficulty level.
Another use case: You have two models on your local device. One is large and fairly powerful but low, the other is smaller, faster and good at tool calls and chat, but not great for writing and reviewing code. If you route between them per request, you can get a better developer experience with preserved performance.
The linked repo aims to help you achieve these things, as do I with the role-model router and protocol that I linked in another comment.
to add to that, for example at the end of implementing a task, where the model runs the formatters, linters, tests, commits, pushes, this could be done by a very cheap model, and only switch to the main model again if something fails hard
there are some cache-busting considerations, but solvable
indeed. i also wrote elsewhere that the current ideal number of models in a pool is probably 2, so if you route between two both will have warm cache, though not the full cache at all times, so you lose a little but not much.
LLMs have no state. There is nothing remembered, nothing new learned. It's the same input , the same output always (unless seeding is randomized). So during a chat it won't matter if every chat turn a different provider is used.
I'm not sure if output of easy commands like "summarize this" are added back to the context? I always assumed they are in a separate UI layer?
There are so many proxies like this now but I can tell you from first hand experience this is not going to work. You cannot just route away from a situation at such a high level especially when we are talking about models that are quite different in behaviour, with different context windows and tuned to different tool uses. The harness is doing all kind of funky things to compensate for issues (like tool call truncation) that a proxy that routes dynamically like this will work against the very same strategies that make the harness work.
Interesting concept, work in theory, but I cannot see this being part of larger system.
This is not choosing between different models, really. You should check the (interesting, yet sadly very slop-padded) readme. It’s about trying to make a binary decision: Is this a hard or easy question, and about making that decision extremely fast. They suggest putting another router that chooses the model behind it. I’m not sure how well it would work, but the idea is interesting and different than other routers.
Has anyone tried the others listed? Any feedback?
We are developing many applications in my company, some of them safety critical. A natural routing way could happen for certain phases of development, and interfaces via git. One agent works on branch a and is responsible for brainstorm planning specs, and the other is responsible code and tests. The first agent creates tickets for the second one and the second one consumes these. This works with today’s standard harness.
Love to see local/cloud routing explicitly supported.
I'm building another router for routing between local and remote models, ShowHN coming up later today. Here's a sneak preview of the github: https://github.com/try-works/role-model
Posted my ShowHN: https://news.ycombinator.com/item?id=48706181
Slight tangent, but “Wayfinder sits behind whatever OpenAI-compatible client you already use” reminds me that descriptions of where proxies sit in the information flow always seem so arbitrary to me:
- “after the client”
- “reverse proxy” (in front of servers)
- “proxy” (in front of client)
"after the client" and "in front of client" can mean the same thing depending on your viewpoint.
Exactly, that’s my point
It'd be nice to just have a command prefix e.g.
/local fix my typo
That’s what I did with Pi, super simple :)
can you send to multiple LLMs to compare responses? From that create a heuristic of which LLM gets what.
I do this manually with a desktop app called BoltAI that lets you continue the whole conversation at your LLM of choice.
This is the way!
I like to think so!
axa
wfdwcZz