🚀 Big shoutout to @Gradient_HQ for making Qwen 235B run on Parallax — the sovereign AI OS we all deserve.
Powerful and built for everyone.
Watch how they did it in minutes 👇
We’re now hosting the @OpenAI gpt-oss 120B model on the Parallax Playground — powered by a RTX 4090 and an Apple M4 Pro, in different part of US.
In the future, everyone will be able to host locally, or with anyone else around the world.
More exciting stuff coming.
Finally, an open model from OpenAI.
gpt-oss-120b is now live on Parallax in Hybrid Mode.
Try it with reasoning, served peer-to-peer across a mesh of everyday hardware.
Had a blast speaking at @BerkeleyRDI and @StanfordSBA about our work on decentralized AI.
Grateful for the brilliant minds, thoughtful questions, and the energy in the room.
Onward 🧠⚡
People always ask me, what’s the point of having consumer-level GPUs host those models, or even doing decentralized AI generally.
I always say:
- Because intelligence shouldn’t be gated.
- Because the future of AI shouldn’t be owned by a few data centers.
- Because there are
Intelligence has been locked in walled gardens. Today, we’re opening the gates.
Parallax now runs in Hybrid mode, with Macs and GPUs serving large models together in a truly distributed framework.
🇰🇷 In Seoul for KBW2025?
Bookmark this thread for Gradient’s full IRL lineup: flagship events, panels, mixers, and more.
Full details and RSVP links below.
Reinforcement Learning is the future tense of intelligence. Echo is how it scales.
Echo is Gradient’s distributed RL framework, running on everyday consumer devices.
From its early experiments, Echo powered a 30B Sokoban model that outperformed DeepSeek-R1 and GPT-OSS-120B.
I’m often struck by how human self-development mirrors Reinforcement Learning.
With strong reward signals from parents, friends, or bosses, we learn fast. With punishment, we avoid mistakes. Sometimes we get stuck in local optima—chasing short-term wins while missing rewards
for self-host, check out this repo we built :) github.com/GradientHQ/par…. It splits LLMs into few chunk of layers and let you host them across your laptops. Mac+Nvidia supported
for self-host, Check out this repo we built :) github.com/GradientHQ/par…. It splits LLMs into few chunk of layers and let you host them across your laptops. Mac+Nvidia supported
Yassssss! for self-host, Check out this repo we built :) github.com/GradientHQ/par…. It splits LLMs into few chunk of layers and let you host them across your laptops. Mac+Nvidia supported
for self-host, you should check out this repo we built :) github.com/GradientHQ/par…. It splits LLMs into few chunk of layers and let you host them across your laptops. Mac+Nvidia supported