webpad
Launch a token
← originals
hosted on webpad gpus · free for $COMPUTE holders

open models

Selected open models from Hugging Face, running on Webpad computers. We cover the compute — holders chat for free instead of paying per request.

all models link to their hugging face repo

MODELS0 / 2 computers ready · 0 models serving
Llama 3.1 8B installing
Meta · 8B◆ Chat
Llama 3.1 8Bmeta-llama/Llama-3.1-8B-Instruct

Meta’s all-round chat model. Fast, smart, great for everyday questions and writing.

Chat free →
Qwen 2.5 7B installing
Alibaba · 7B◆ Chat
Qwen 2.5 7BQwen/Qwen2.5-7B-Instruct

Strong multilingual chat model with long context. Great at following instructions.

Chat free →
Mistral 7B installing
Mistral AI · 7B◆ Chat
Mistral 7Bmistralai/Mistral-7B-Instruct-v0.3

The classic efficient open model. Quick answers, low latency.

Chat free →
DeepSeek R1 8B installing
DeepSeek · 8B◆ Reasoning
DeepSeek R1 8Bdeepseek-ai/DeepSeek-R1-Distill-Llama-8B

Reasoning model that thinks step by step before answering. Best for math, logic and hard questions.

Chat free →
Qwen 2.5 Coder 7B installing
Alibaba · 7B◆ Coding
Qwen 2.5 Coder 7BQwen/Qwen2.5-Coder-7B-Instruct

Code specialist: writes, explains and fixes code in dozens of languages.

Chat free →
Llama 3.2 Vision 11B installing
Meta · 11B◆ Vision
Llama 3.2 Vision 11Bmeta-llama/Llama-3.2-11B-Vision-Instruct

Sees images: describe a picture, read text in it, answer questions about it.

Chat free →
COMING NEXT
Image generation

Stable Diffusion & FLUX image models are being installed on the computers.

Audio

Whisper speech-to-text and open voice models are being prepared.

01Pick a model
02Connect a wallet holding $COMPUTE
03Chat — the computer answers
04No per-request bill, ever

$COMPUTE · GWe3TNg1UPxeaSazxK8U49PDAqXCaX9Jc8zbttFLpump