Show HN: Distill and serve models with frontier quality for half the cost

https://github.com/experientiallabs/world-model-optimizer
Hi HN, we built world-model-optimizer, an open source tool to continually improve models specialized to agents.

Agent traces you already capture are opportunities to get signal on how to make your model cheaper, faster, better.

We do this by continuously a) distilling relevant chain of thought from larger open source models into smaller ones, b) model routing to frontier + OS models, and c) token compaction to remove noise and save on tokens.

`wmo build` allows you to build a simulation to optimize against with your agent traces with your OpenRouter key

`wmo optimize` trains a router, compaction, and distills chain of thought from a larger model into your specialized model

`wmo serve` gives you an endpoint for your model

When you call your model, behind the scenes a router decides which tasks should go to the frontier versus your model. Tinker continually trains as new traces arrive.

We're also working on a hosted solution that does continual training + serving for you https://experientiallabs.ai

Comments

Art9681Jul 27, 2026, 12:15 AM
The absolute best way to prove this works is by releasing a model that was fine-tuned with this method and then showing benchmarks depicting the improvement delta between the base model and the fine tuned one.

The work is not done. Then release it to the masses and wait a few days for the actual real world anecdotes.

Until then, this is noise.

ReubendJul 27, 2026, 12:19 AM
Yeah, this is just slop. No benchmarks, no concrete case studies, just some vibecoded "platform" to finetune models on your own traces.

Which is an idea that has some value, but also some weaknesses. And this implementation of it isn't forthcoming with that concept. You have to really dig in to understand what they're even talking about.

SilenNJul 27, 2026, 12:25 AM
Happy to answer any qs.
SilenNJul 27, 2026, 12:22 AM
Valid criticism. Happy to answer any qs. We're still working on solidfying results.
dangJul 27, 2026, 12:24 AM
Ok, I think it is in your interest to wait until you have more to show, and we'll be happy to help you with reposting it once it's ready.

Waitlists are against the Show HN rules (https://news.ycombinator.com/showhn.html), and you're likely to get a lot of community pushback if you post before there's enough substance for users to sink their teeth into.

SilenNJul 27, 2026, 12:30 AM
Thanks for the heads up, removed mention!
irishcoffeeJul 27, 2026, 12:20 AM
Benchmarks are the ultimate consolidation of halnons razor.
teravorJul 27, 2026, 12:21 AM
[flagged]
dangJul 27, 2026, 12:22 AM
"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."

https://news.ycombinator.com/newsguidelines.html

adriancoJul 27, 2026, 12:08 AM
Local models need to be tuned to work well so this looks useful. Seems to be for general purpose model serving. I’ve been using https://github.com/adrianco/retort to run experiments for coding models across 13 different programming languages to see which frontier and local models work.
SilenNJul 27, 2026, 12:14 AM
That's cool, thanks for sharing!
jack_ppJul 27, 2026, 12:05 AM
Not sure I get it. The model you're improving is local? If so how do you even calculate cost compared to an API
SilenNJul 27, 2026, 12:21 AM
Open source models.

wmo routes requests between frontier models and open source models that continuously train using Tinker. As the smaller models improve, more traffic gets routed to them.

Calculating cost is just tokens in/out.

digitaltreesJul 27, 2026, 12:07 AM
Cool project
SilenNJul 27, 2026, 12:31 AM
Thanks :)
surroundJul 27, 2026, 12:26 AM
The title is misleading. This is model routing, not distillation.
SilenNJul 27, 2026, 12:31 AM
Fixed formatting which will help with readability. We do routing, distillation, and token compaction.
yiyingzhangJul 27, 2026, 12:07 AM
Cool idea! How do you guarantee privacy?
SilenNJul 27, 2026, 12:22 AM
It's open source!

We do have a platform we'll be launching as well to manage training + serving for you which will require more diligent privacy guarantees.

rgloverJul 27, 2026, 12:03 AM
Excited to play with this.
SilenNJul 27, 2026, 12:31 AM
Let me know if you have any questions!
renezander030Jul 27, 2026, 8:10 AM
[flagged]