Show HN: sllm – Split a GPU node with other developers, unlimited tokens
Scored daily by a customisable AI persona to surface the most relevant engineering leadership news.
sllm offers GPU node sharing for LLMs, directly addressing cost and access in AI infrastructure.
GPU-sharing platform sllm shut down after its cohort-based activation model never reached the required threshold, meaning no users were billed and no API usage occurred. The team will delete all user data within 30 days and recommends alternatives like Together, Fireworks, and Groq for pay-as-you-go access to models like Qwen and Llama.
- Evaluate Together, Fireworks, and Groq as drop-in replacements for sllm's planned GPU-sharing service, especially for high-throughput access to Qwen and Llama models.
As a solutions architect evaluating AI infrastructure, this shutdown highlights the challenges of building GPU-sharing marketplaces and confirms that established pay-as-you-go providers are currently more reliable for deploying open-weight models.