frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Show HN: Echo – Fable-level results at 1/3 the cost using open-weight models

95•adam_rida•1h ago•40 comments

Show HN: Palmier Pro – Open-source macOS video editor built for AI

https://github.com/palmier-io/palmier-pro
74•harrisontin•5h ago•15 comments

Show HN: Trifle – Open-source analytics that stores answers, not events

https://trifle.io/
21•iluzone•1d ago•2 comments

Show HN: OneCLI – OSS credential gateway that keeps secrets out of AI agents

https://github.com/onecli/onecli
58•Jonathanfishner•5h ago•24 comments

Show HN: Claude-thermos – keeps your Claude session warm for you

https://github.com/izeigerman/claude-thermos
39•s0ck_r4w•3h ago•25 comments

Show HN: Remux – an open-source tmux workspace designed for iPhone

https://github.com/h3nock/remux
72•bitwise42•6h ago•29 comments

Show HN: Whetuu – a zero-config cross-shell prompt written in Zig

https://yamafaktory.github.io/whetuu/
42•yamafaktory•6h ago•25 comments

Show HN: macOS menu-bar manager for SSH port forwards

https://github.com/lx2026/RelayBar
35•linxy97•6h ago•5 comments

Show HN: Notebooker.ai – NotebookLM alternative, your own models, keys, storage

https://notebooker.ai/
3•mistakevin•1h ago•0 comments

Show HN: I built a zero-latency developer tool suite in pure Vanilla JavaScript

https://omnideck.cc/
16•AAPD_Studio•6h ago•10 comments

Show HN: Dally – A little every day adds up

https://mctools.site
14•totaldude87•7h ago•1 comments

Show HN: Bento - An entire PowerPoint in one HTML file (edit+view+data+collab)

https://bento.page/slides/
977•starfallg•1d ago•228 comments

Show HN: Hanky – ETL style framework for loading flash cards into Anki

https://github.com/Haeata-Ash/hanky
8•funfruit•7h ago•0 comments

Show HN: GapQuery – find app-market gaps from 35,600 apps and their reviews

https://www.gapquery.com
5•northify•6h ago•0 comments

Show HN: Syncle – Open-source real-time database synchronization engine

https://github.com/osmanahmadxai/SYNCLE
2•osmanahmadxai•3h ago•0 comments

Show HN: Setoku – Self-hosted knowledge server for AI agents

https://setoku.com/
3•rgbrgb•3h ago•0 comments

Show HN: Historic adobe buildings mapped across America

https://adobepueblo.com/
3•not_wowinter13•3h ago•0 comments

Show HN: Flow Matching model inference in C

https://github.com/ecrasevvv/c-fm/tree/master
2•ecrasevvv•3h ago•0 comments

Show HN: I made YAFL – a E2EE file handoff for AI agents

https://yafl.dev
4•marcushorndt•6h ago•1 comments

Show HN: I simulated closing the Strait of Hormuz on real oil trade data

https://globaloilnetwork.staffinganalytics.io/
5•eliotho•8h ago•3 comments

Show HN: What's AI's go-to, public or private healthcare?

https://www.modelbias.ai/prompt/public-or-private-healthcare
8•bnfcl•6h ago•3 comments

Show HN: Cactus Hybrid: We taught Gemma 4 to know when it's wrong

https://github.com/cactus-compute/cactus-hybrid
183•HenryNdubuaku•1d ago•40 comments

Show HN: NeoBox – a native macOS NeoGeo emulator front end

https://neo-box.app
4•harisamin•5h ago•0 comments

Show HN: DeepSQL – A self-hostable DBA agent for Postgres and MySQL

https://deepsql.ai/
50•venkat971•3d ago•29 comments

Show HN: Web swing through midtown NYC

https://www.swingnyc.com/
75•shahahmed•1d ago•26 comments

Show HN: Find undervalued bikes near you with ML

https://bike.broker
3•azuroscloudapi•5h ago•0 comments

Show HN: A primer for non-coders to design software via structured AI dialogue

https://github.com/CoreGrowthLabs/CoreGrowthPrompting/blob/main/README.en.md
5•CGP_Labs•5h ago•1 comments

Show HN: MinkNote – Native macOS notes and journaling using Markdown files

https://muse23.com/apps/minknote/
2•demianturner•5h ago•0 comments

Show HN: ManiDesk – SQL workspace joining market, macro and fundamental analysis

https://mani-desk.com/
3•franklin_m•5h ago•2 comments

Show HN: HN Hall of Fame – browse 3,100 legendary Hacker News links

https://www.orangecrumbs.com/hall/
204•oyster143•1d ago•38 comments
Open in hackernews

Show HN: Echo – Fable-level results at 1/3 the cost using open-weight models

95•adam_rida•1h ago
I’ve been building Echo (https://echo.tracerml.ai/), an experiment in making one AI system out of a pool of open-weight models rather than choosing a single model and using it for every task.

It started with a simple experiment. I took a group of models, including GLM-5.2, Kimi K2.7 and others, and ran them on the same evaluations. Then I measured what would happen if, for each problem, you somehow knew in advance which models would be useful and how their outputs should be combined.

That hypothetical system performed substantially better than any individual model in the pool. Of course, it is not something you can actually deploy because it relies on knowing which decisions were good after seeing the result. Echo is my attempt to recover some of that advantage without having that information in advance.

For each request, Echo decides how much computation to allocate, which models should participate, and how their work should be combined. Some prompts may only need a relatively small amount of inference, while others benefit from multiple models working on different parts of the problem.

One thing that surprised me while building it was how complementary the models are. A model that is clearly weaker overall can still be extremely useful on particular problems or as part of a combination.

On my first evaluation mix, Echo consistently performed better than the best individual model in its pool. It also reached roughly the same aggregate result as Fable, which I used as one of the stronger comparison systems, at around one third of the inference cost.

There are still some cases where Echo makes the wrong allocation or combination decision. I’m currently spending a lot of time understanding those failures, as well as testing whether the same approach holds up on coding and agentic tasks where measuring the quality of each decision becomes much harder.

I built a chat interface (echo.tracerml.ai) and an OpenAI-compatible API (https://echo.tracerml.ai/docs/api) so the system can be tested outside the evaluation setup.

Here is a short/high level video on how it works: https://www.youtube.com/watch?v=lJFJSvOdXhg

I wrote up the evaluation methodology, individual model results, costs and current limitations here: https://echo.tracerml.ai/eval

I would love for you to try it! Especially if you hit any weird failure cases or places where the allocation looks unintuitive.

Comments

kamranjon•1h ago
No benchmarks, no info on which models are used, ai generated video, just a signup page with nothing else.

Anyhow, this kinda reminds me of that quote about architecture: "We replaced our monolith with micro services so that every outage could be more like a murder mystery."

guessmyname•1h ago
> No benchmarks, no info on which models are used, […]

The benchmarks are here → https://echo.tracerml.ai/eval/

They are not good benchmarks but at least they exist.

seizethecheese•24m ago
I've been working on a similar project and I found that it's easy to replicate Fable results if you use saturated benchmarks.

In my project, I wasted a huge amount of time trying to improve GPQA Diamond results above ~93% range. I realized my mistake when Fable dropped and made no improvement on this benchmark vs. Opus.

j45•1h ago
It's easy enough to copy and paste in a prompt, no?

Eval tests while giving general indicators might not be similar for each use case.

adam_rida•51m ago
The evaluator is public here: https://echo.tracerml.ai/eval/

It currently exposes 907 stored rows across seven benchmark families, with prompts, outputs, grades, and cost records. More benchmarks are coming soon.

Echo does not disclose its per-request routing decision because that policy is the product. We can, however, publish some of the eligible open-weight model pool, version dates, aggregate allocation mix, and evaluation settings without exposing the request-level recipe.

New video is also being made.

codekansas•42m ago
It's basically trying to replicate OpenRouter, which works pretty well and has a lot of nice features to abstract away any single provider, such as failover, metering, autoswitching, etc. It's actually a really smart infrastructure abstraction.

I just wish this were solving an actual problem rather than being a fairly transparent attempt to say something approximating, "Hey VCs, OpenRouter just became a unicorn but I can basically vibe code it"

Calling it "Fable-level" feels intellectually lazy / dishonest, but then again, what do you expect when there's so much money on the table.

seizethecheese•1m ago
OpenRouter's model router for coding isn't really sophisticated. This is very different.
tj800x•1h ago
No single signin. Privacy policy allows training. No try it first without credit card. It's a good idea, but this looks premature.
adam_rida•54m ago
You are right that the privacy wording was too broad. We are fixing it now so it states explicitly that Echo does not use customer prompts, files, chats, or outputs to train or fine-tune models. We are updating the matching Terms language at the same time.

Echo also starts users with free credit and does not require a credit card to try it. The current signup flow did not make that clear enough, so we are fixing that presentation too.

Thanks for calling both out.

cynicalsecurity•19m ago
I assume trying without a credit card will bankrupt him immediately.
fneddy•1h ago
That’s basically the same idea IBM advertises with Bob?
jmaw•1h ago
I think approaches like this have potential. Only time will tell. This reminds me of the mixture of experts taken by deepseek r2 (I think it was r2, at least), but less specific models I guess.

I have often wondered how tools like GHCP choose the best model for the job when set to "auto".

Alifatisk•1h ago
This reminds me on OpenRouters report that combining multiple different models gave comparable performance to Fable 5. I think this approach has lots of potential. Maybe OpenAi was ahead of its time with GPT-5 (it being a router to different models rather than just being one new model)
seizethecheese•28m ago
You might be interested in a project that I'm working on, which is kind of like OpenRouter Fusion, but instead of waiting for all models to synthesize, we stream the best model immediately and background the rest. The background models then reconcile with an emoji reaction and optional reply. It gets similar results to Fusion and is a lot faster! There's a free version that leverages open weight models here: http://pellmell.ai.
j45•59m ago
If you copy Perplexity, they let you have the first few rounds of chat for free to get you going before asking to sign up.
codekansas•54m ago
Fable-level, yea, but can it run gstack?
jambalaya8•53m ago
Might want to rethink the name to avoid an Amazon issue.
wizche•51m ago
how does this differs from OpenRouter fusion?
cawingcrow•38m ago
how does this differs from Sakana Fugu?
adam_rida•23m ago
Fugu includes closed source models in its pool
bnjemian•47m ago
I don’t find the recent spate of blog posts and systems delegating and combining LLMs to get better performance particularly interesting. Especially given that anyone who’s taken an ML 101 course has learned about ensemble methods.

While an LLM isn’t what you’d traditionally consider a weak learner, the theorems on learning systems clearly point to them being so in this context. The feigned surprise at combining them to yield better results seems disingenuous.

Even so, the work to predict which models are best suited for which task, how to delegate, and how to combine their outputs is interesting, especially if you’re placing a cost minimization objective on it. That said, this isn’t too far off from what many AI labs are already doing.

abernard1•29m ago
The entirety of "agents" and tool calls is a process of combining LLMs to get better results. Is it the same LLM in many cases? Yes. But it doesn't have to be.

It's the natural move that happened after people realized you couldn't throw away half a century of AI research.

Most of these focus on costs. But it is simply the case that the one-shot output did not scale for harder problems on workflows.

ototot•37m ago
Is this yet another Sakana Fugu / OpenRouter Fusion?
kachnuv_ocasek•7m ago
Looks like it but with open-weights only.
yonatan8070•34m ago
I'm not an expert on this, but this sounds a lot like a larger-scale MoE (Mixture of Experts) type of architecture.

As I understand it, in an MoE model, you essentially have hundreds of smaller sub-models ("experts") that are good at different tasks, and for every generated token, a single "master" model chooses which ones are most relevant to participate, and you only activate them.

janalsncm•31m ago
In MoE systems the routing decision is made per-token, not per prompt or task. It’s one of ML’s many confusing naming conventions.

Even more confusingly, there are older pre-LLM MoE systems which ensemble and pool the predictions from multiple sub-components. For example in a random forest you could take the majority vote of the decision trees or the average of their numerical predictions.

After that, we developed neural net architectures for predicting a single thing like whether the user will click on your ad. An MMoE is in the same family.

And so now we are at massive MoE networks for LLMs which have similarities with MMoE in that the “decision” is about the very next token to predict.

bbstats•23m ago
M-o-MoE
maxdo•19m ago
such a scam, there is only one fable-like model, that somewhat behind, it cost half, not 3x. so from here you can stop reading.
janalsncm•19m ago
Intuitively, your savings depend heavily on how hard the tasks are in the first place. If you have a base rate where 99% of your tasks can be routed to a cheap model, yeah, you can save a ton by not using Fable for that.

So “1/3 the cost” really depends.

ninjahawk1•17m ago
I’m very confused on what this is, my initial thought was “oh nice, open source router.”

I go to the website…and it’s a sign up. I expected a repo. Otherwise how do I use it? As a SaaS? Yeah right.

Oh well I guess at least the benchmarks are good…I find the benchmarks and many are either not present or are not what the title claims.

My main question is how this has so many updoots from HN, probably the passerby not looking closer for sure.

I mean no offense and I really do wish you best on this, but it seems like what we used to call back in the day, vaporware.

jacobgold•14m ago
> Fable-level results at 1/3 the cost using open-weight models

But we get ~$2500/mo worth of Fable credits for $200/mo on Anthropic pan? I'm still confused why people (who don't have to use API billing) are chasing open weight models based on cost.

teruakohatu•12m ago
Because that is a short term solution, it won’t be offered forever. Large organisations have to purchase credits at $/tokens. Eventually everyone else will too.
sscaryterry•4m ago
This is what OpenAI and Anthropic are trying to make everyone believe. Most accountants will flinch at this (they already are).

The $200 odd plans are already out of reach of many, many people.

The attrition of customers if they were to get rid of these subscriptions plans would be untenable.

dberg•8m ago
doesnt exist for enterprise plans
combyn8tor•3m ago
You may want to consider the incomes of developers outside the US, students, unemployed. $200/month is a lot to a lot of people.
dbbk•1m ago
They can still write code.
gruez•30s ago
The point still stands. The chinese labs don't have super discounted plans, so if the price per task benchmarks[1] are correct, and we apply the discount, you'll actually be paying more by using cheaper chinese models and this technique.

https://artificialanalysis.ai/agents/coding-agents#artificia...

recursivegirth•2m ago
Enterprise plans pay API costs, they're only subsidizing individual accounts and that's only because they have to compete with open-source models.

When they are successful at making those illegal/inaccessible, both Anthropic and ChatGPT are going to rip the band-aid off and inference will only be sold to those that can afford it.

The good ole American way.

dluan•2m ago
So this is the dogpile.com of the askjeeves, alta vista, and lycos approach? Time is a flat circle?