Pareto Inference (@precursorlabs) header
Pareto Inference (@precursorlabs) profile picture

Pareto Inference

@precursorlabs

optimal performance and unbeatable prices trusted by @vercel @opencode @commandcodeai

 NYC
paretoinference.com/
Joined April 2021
  • Tweets12
  • Following5
  • Followers4.7K
4Media
Photo by Pareto Inference (@precursorlabs)Photo by Pareto Inference (@precursorlabs)Photo by Pareto Inference (@precursorlabs)Photo by Pareto Inference (@precursorlabs)
Pareto Inference retweeted
Jakub Janiak (@oogway_defi) profile picture
Jakub Janiak
@oogway_defiOct 13
developers love pareto inference
Image from Jakub Janiak (@oogway_defi)
1
3
1
13
882
Pareto now offers zero data retention: we don't keep your prompts or outputs after a request finishes. paretoinference.com/zdr
Image from Pareto Inference (@precursorlabs)
1
2
5
825
Early look at continually trained, local decision models. The model that makes the same decision you would is the most aligned.
hands (@handsdiff) profile picturehands
@handsdiffSep 21
Just dropped a preview of some of the latest work I'm doing @precursorlabs to distill personal judgment into local models continuously. Our main finding is that models are overall pretty bad at predicting the decisions you actually make on a day to day basis. If you had a local "decision model" that was able to actually make the decision you would, given the same context, you could move 1000x faster. There seems to be consistent learning as a function of data, pointing to something akin to local scaling laws!
Image from hands (@handsdiff)
8
733
$3/week for GLM and DeepSeek API access Pareto Pass is officially live 🎉 70 tok/s median on GLM-5.3 Join with a referral: $1 for your first week Refer a friend: earn $3 subscription credit after their first $3 renewal Get started: paretoinference.com
🤖 Made with AI
Image from Pareto Inference (@precursorlabs)
4
4
2
16
50,915
Pareto Inference retweeted
Alex Janiak (@0xjaniak) profile picture
Alex Janiak
@0xjaniakJul 28
After digging into how inference providers cut corners, I tested GLM 5.2 on FP4 endpoints. They charge less per token and more per answer. A listed 32% discount became an 11% premium per solved problem. Price parity became a 40% premium. Have a read!
Image from Alex Janiak (@0xjaniak)
2
2
2
11
968
LLM interaction today is slow and clumsy. We are exploring solutions to this.
hands (@handsdiff) profile picturehands
@handsdiffJul 22
Context costs 10x more than tokens

Context costs 10x more than tokens

Article · @handsdiff
10
1,486
First of our notes on the AI buildout: we went down a rabbit hole on "black box" inference APIs
Alex Janiak (@0xjaniak) profile pictureAlex Janiak
@0xjaniakJul 15
Another post! An analysis of inference market mechanics and opaque provider practices. “making a model open is only half the battle.”
Image from Alex Janiak (@0xjaniak)
1
5
1,079
Introducing Precursor Labs: a research company studying the infrastructure for collective intelligence. The next generation of autonomous systems will need to coordinate, reason under uncertainty, and allocate resources in dynamic environments. We're working on how. ↓
Image from Pareto Inference (@precursorlabs)
5
8
2
71
19,277
We're a small team of engineers, financiers, and company builders working across three threads: multi-agent systems, the AI buildout and market structure, and decision-making infrastructure. We'll be sharing research and essays here and on our Substack (precursorlabs.substack.com/). Follow along!
1
13
1,290
Load more