Overview
Strengths & Best For
Lyceum is an academic GPU cloud offering H100 and A100 instances with research-friendly policies, flexible on-demand pricing, and configurations tailored for AI experimentation and academic research teams. Accessible pricing and a focus on the research community make it a practical option for universities and independent researchers who need GPU compute for LLM experimentation and model training without enterprise overhead. A strong choice for academic AI teams that want affordable, flexible GPU access.
- Research-friendly
- Flexible pricing
- Academic focus
- Very small provider — limited scale
- No enterprise SLAs
- Limited documentation and support resources
No reviews yet — add yours ↗
Runcrate is a GPU cloud founded in 2024 offering on-demand instances with per-second billing and no minimum commitment — the meter stops the moment you stop the instance. The platform aggregates GPU capacity across eight global regions spanning North America, Europe, and Asia, covering 14 GPU models from RTX A6000 and A100 through H100 and RTX PRO 6000. Full root access via SSH, Docker support, and custom images are included on every instance. An OpenAI-compatible inference API serves open-source models billed per token, and reserved or dedicated capacity is available for teams with monthly volume commitments.
- Per-second billing — no minimum commitment
- OpenAI-compatible inference API
- 14 GPU models including H100 and RTX PRO 6000
- 8 global regions across NA, EU, and Asia
- Full root SSH access with Docker and custom images
- Reserved and dedicated capacity available
- No enterprise SLAs or uptime guarantees
- Smaller provider — limited scale vs hyperscalers
- No managed platform or containers
No reviews yet — add yours ↗
Live GPU Pricing
Region Coverage
Popular Comparisons
Provider pages
Key differences
Lyceum
GPU cloud for researchers
Runcrate
GPU cloud with per-second billing and an OpenAI-compatible inference API