Kled AI

The first consumer data marketplace. Sourcing the largest licensable multimodal datasets on the planet.

[Kled AI @useKled]

This is Kled V3. We've solved data collection for artificial general intelligence.

Any consumer dataset that can exist, can now be collected from physical reality in under 72 hours.

All powered by the largest and most comprehensive data application layer on the planet. (Thread)

[Kled AI @useKled]

Kled does not sell company data, we only sell consumer data via our data collection app.

Company data brokers do not eat our margins or anything else.

All competition is welcome in consumer and robotics as well, just don't lie to labs or companies that hurts everyone in this space. - Kled team

[Kled AI @useKled]

@avipat_ we are laser focused on becoming the largest gig app in the world.

[Kled AI @useKled]

@avipat_ https://t.co/etm0tdDGjo

[Kled AI @useKled]

@0xDamien we are paying double for the next few weeks on this task, rates will be disclosed publicly soon for now the task is labeled as the highest pay and found at the top of the task feed in app

[Kled AI @useKled]

Labs with a custom spec can explore the Kled marketplace and submit a request today.

Our team reviews every submission and follows up directly with qualified labs: https://t.co/etm0tdEe8W

[Kled AI @useKled]

Over the weekend, Kled received a complex multichannel Korean audio task from one of the frontier labs. Within 48 hours, it was live in the Kled app.

We are now sourcing thousands of Korean speakers to produce millions of hours of multichannel audio.

The spec calls for per-speaker isolated channels, speaker attribution, persistent speaker IDs across recordings, a five-hour capture limit per speaker, 16 kHz+ audio, and documented consent for AI training use.

Available today on iOS: https://t.co/2pLpoRGXtK

[Kled AI @useKled]

the kled marketplace is now live: https://t.co/etm0tdEe8W

labs can request and browse datasets from the kled contributor network and a curated network of enterprise partners.

[Kled AI @useKled]

Introducing Kled Labs.

Every major AI model learned from the internet. We're betting on people instead.

One goal: train our own models to beat closed models at important human tasks, using our data alone.

All data is sourced from our data application layer and our network of 600,000+ contributors.

We're hiring: https://t.co/jCggnuvaX3

[Kled AI @useKled]

@avipat_ if you are a lab that is not already in our network, you can dm us for early access

[Kled AI @useKled]

@avipat_ https://t.co/BTsi6tTdNT

[Avi Patel @avipat_]

https://t.co/cAMMf9fY9J

[Kled AI @useKled]

@saintofthehour @andyfolio @foliostudios the best

[Avi Patel @avipat_]

say hi to kled v3, the data collection layer for agi.

any dataset a lab can describe, collected from the real world by 500,000 people in under 72 hours.

[Kled AI @useKled]

@andyfolio @foliostudios @avipat_ @saintofthehour the best out there

[Kled AI @useKled]

@avipat_ patel mode is back

[Kled AI @useKled]

The latest version of Kled is available today on the iOS App Store. AI labs looking for structured data can reach out or submit a dataset request at https://t.co/E5QrrdmlM2.

[Kled AI @useKled]

This is Kled V3. We've solved data collection for artificial general intelligence.

Any consumer dataset that can exist, can now be collected from physical reality in under 72 hours.

All powered by the largest and most comprehensive data application layer on the planet. (Thread)

[Kled AI @useKled]

Kled's previous home screen was built around a single upload button for photos already in a contributor's camera roll. Structured tasks lived in a separate tab that most contributors rarely opened.

This meant we could only collect the data people already had, not the data labs actually needed.

V3 replaces the home screen with a task discovery feed. Every live task is ranked for each contributor, and lab-priority tasks appear first with their pay shown upfront, so the work that matters most to the frontier is what 500,000 contributors see first.

The feed is powered by a recommendation system that learns where each contributor performs best, measured by acceptance rate across modality, task type and region, and personalizes their feed around it. Contributors can also adjust it themselves. Better matching leads to higher acceptance, higher earnings and higher-quality data.

Supply is now programmable. If a lab needs 40,000 minutes of egocentric video from Southeast Asia this week, we route contributors to it directly instead of waiting for the network to produce it.