Raxi Labs

Raxi Labs builds open-source frameworks for LLM systems, created by @kylercph

[Raxi Labs @RaxiLabs]

We just opened the Raxi Labs Discord.

If you care about:
- multi-LLM systems
- model verification
- evaluation and arbitration
- open-source AI infrastructure
- what comes after Raxi Court

join us here👇
https://t.co/0IFWnYa2DC

[Raxi Labs @RaxiLabs]

@beyoumf Of course

[Raxi Labs @RaxiLabs]

Raxi Court turns that manual workflow into infrastructure: parallel generation, anonymized comparison, arbiter scoring, and stored reports.

[Raxi Labs @RaxiLabs]

@lochan_twt Tech is not dead

[Raxi Labs @RaxiLabs]

Raxi Court does not stop at a second opinion.

It uses multiple candidate responses, role-based arbiters, weighted scoring, and reportable evidence for the final verdict.

[Raxi Labs @RaxiLabs]

@elonmusk wow; glad to see starlink be integrated into something!

[Raxi Labs @RaxiLabs]

Raxi Court is less interested in model leaderboard theater and more interested in whether multiple answers survive anonymous arbitration.

[Raxi Labs @RaxiLabs]

@kylercph @ellawurld0 That was my personal account speaking, sorry about that

[Raxi Labs @RaxiLabs]

We just opened the Raxi Labs Discord.

If you care about:
- multi-LLM systems
- model verification
- evaluation and arbitration
- open-source AI infrastructure
- what comes after Raxi Court

join us here👇
https://t.co/0IFWnYa2DC

[Raxi Labs @RaxiLabs]

@Tekeee Thank you for noticing us @Tekeee

[Raxi Labs @RaxiLabs]

@Zinny_Edmund No, not really.

[Raxi Labs @RaxiLabs]

Want less model bias in evaluation?

Stop telling the judges which model wrote what.

Raxi Court shuffles and anonymizes responses before arbitration.

[Raxi Labs @RaxiLabs]

@akinkunmi Ur damn right about that

[Raxi Labs @RaxiLabs]

@petaldairies Thank you sincerely

[Raxi Labs @RaxiLabs]

@ritu_twts We don’t vibe code our tools

[Raxi Labs @RaxiLabs]

@itsaaroshi Python 🐍❤️

[Raxi Labs @RaxiLabs]

We are building Raxi Labs in public.

That means the ideas, mistakes, revisions, and frameworks all happen in the open.

View our GitHub repo here: https://t.co/zjqq79aU4t

[Raxi Labs @RaxiLabs]

@ube_codes no; we are in for improving the ai space @ube_codes.

[Raxi Labs @RaxiLabs]

@fidexcode 0, ours is actually human written for once.

[Raxi Labs @RaxiLabs]

@abhijitwt Hey @abhijitwt, here's mine:

https://t.co/zjqq79aU4t

[Raxi Labs @RaxiLabs]

Raxi Court runs multiple models on one prompt, anonymizes their responses, scores them with separate arbiters, and saves a traceable report.

That is the kind of AI infrastructure that matters.

[Raxi Labs @RaxiLabs]

We just dropped the latest (and cleanest) README banner in our project history. Check it out here:

https://t.co/zjqq79aU4t