AI Developer Cloud. AI infrastructure developers trust. discord.gg/runpod
- Tweets 811
- Following 17
- Followers 9,058
- Likes 857
In 2023, everyone was talking about training. That conversation still matters, but it's not what's eating your week anymore. The fight moved to inference, latency per token, throughput under real load, and a GPU bill that grows faster than you'd like...
7
1
7
2,680
A thousand hours of audio costs $7 to transcribe on the right GPU. $507 on the wrong one.
We benchmarked Whisper large-v3-turbo across 23 GPUs and the results aren't what most people expect.
Read the entire breakdown here:
runpod.io/articles/guides/be…
1
4
286
Always humbled and excited to see what our customers run on Runpod
We now run the largest frontier lab composed entirely of autonomous AI researchers.
Introducing Primus Society: a society of agents composed of thousands of researchers working under structured institutions designed to solve the world’s toughest problems. We believe this is how AI research should run at scale, safely and productively: an entire society of agents with institutions and purpose.
One discovery we can already share is that the society has discovered a novel result which improves model training by 30%.
Primus Society was inspired by the structures that have organized science for centuries and stress-tested against what’s known about how populations of AI agents fail.
Everything is observable. You can open the virtual city in a browser and read what any researcher is working on.
Learn more about Primus Society here:
lab.cloud/society
The biggest opening in the AI race is running the largest well-governed organization of AI researchers in the world. We’re building the institutions that let a million AI scientists safely tackle the world's toughest problems in AI and beyond. And we’re doing it right here in Canada.
1
9
1,927
You shouldn’t have to choose between the cloud your dev team already uses and the cloud your company can approve.
That’s the idea behind Runpod Enterprise. Start self-serve, prove the workload, and formalize it under an enterprise agreement when you’re ready.
The platform stays the same, but the capacity, controls, and support grow with you.
If you’re ready for that conversation, here’s where you can learn more:
runpod.io/blog/runpod-enterp…
1
3
335
We're going live tomorrow with Greg Wester and Vijay Chauhan.
In 30 minutes, they'll take you through what separates enterprise-grade AI compute from enterprise-priced AI compute.
We're hosting two sessions (6 AM & 11 AM PT), so just pick whichever one works best for you!
Register here: resources.runpod.io/enterpri…
314
We’re co-hosting a meetup together with Flox & Apify. Demos, drinks, bites, and LEGO (for all the builders out there).
Join us at Stage 9, Room 211B at 6:30 PM on Thursday.
You can register here: luma.com/apify-qplc
318
"If you simply drop a cutting-edge LLM into the average corporate data warehouse, it doesn't become a brilliant data analyst. It becomes a confident idiot."
Our Head of Data, Charlotte Daniels, shares what building an internal AI agent at Runpod taught us about data foundations, schema design, and why the model is never the bottleneck.
Read the full piece on InfoWorld:
infoworld.com/article/422297…
1
6
667
Nearly three years ago, Rendair started with just one serverless endpoint on Runpod.
They now run 22 in production, which have generated over 13 million images.
See how they built it in our latest case study:
runpod.io/case-studies/how-r…
10
539
We've shipped a lot of new stuff lately, and we want to show you.
On September 24, Greg Wester is walking through the latest Runpod updates live.
You can register through the link below:
resources.runpod.io/enterpri…
1
5
462
Browser-use agents that read the DOM don't need a vision model at all. Screenshot-based agents need one running for every single action.
The GPU you should choose depends entirely on which architecture you picked.
We mapped it out:
runpod.io/articles/guides/gp…
2
1
10
744
Which GPU should you use for embedding workloads?
We benchmarked seven models across 24 GPU types and three serving engines to find out.
See the article below for the full breakdown:
runpod.io/blog/gpu-embedding…
1
1
3
361
Did you know that you can self-host n8n on Runpod?
Our team wrote a guide on how to set it up:
runpod.io/articles/guides/se…
4
390
We’ll be at @WeAreDevs in San José next week.
And on September 25, our own Jessica Garson Beauchemin will show you what it’s like to manage GPU infrastructure through MCP instead of a console.
See you at Stage 5, 11:00 AM PT 😄.
1
2
336
You don’t get to plan when a Product Hunt launch goes viral.
But that’s what happened with Coframe when their Living Images product hit #1.
Read more about their story here:
runpod.io/case-studies/cofra…
1
306
Today we’re launching Global Volumes in beta.
Now you can mount elastic, region-independent storage into a Pod in any Runpod data center.
Just store a model once, deploy your Pod where the GPUs are available, and access the same files at /workspace-global.
See how you can get started here:
runpod.io/blog/global-volume…
7
1
13
698
Runpod retweeted
We’ll be at the @runpod booth at @TheLeadDev NYC. Come say hi of you are around!
ALT Woman at Runpod booth
1
7
364
Asking yourself whether to eat the cost of your Kubernetes GPU stack?
We put together an honest look at what the retrofit actually costs and when a GPU-native approach makes more sense.
runpod.io/articles/guides/gp…
2
4
1,274
The sales process covers pricing, specs, and roadmap.
Nobody talks about what happens six months in when capacity gets tight, or something goes wrong at 2 a.m.
Join our webinar on September 24, where Greg Vijay will discuss what enterprise-grade should actually get you after the contract is signed.
resources.runpod.io/enterpri…
1
3
530
Adding GPUs doesn't automatically mean more throughput.
NVLink vs PCIe is a 14x bandwidth gap, and NCCL picking the wrong network interface can silently cut your training speed in half.
We wrote the full multi-node cluster guide.
runpod.io/articles/guides/gp…
1
2
4
539
Already know a workload can run and need predictable access to the GPUs powering it?
A private GPU pool is often the best fit.
But we kept hearing the same questions come up from customers about these pools.
We just wrote a guide to answer them:
runpod.io/blog/private-gpu-p…
6
493