# ClearCompute > ClearCompute is the decision layer for inference. It narrows every model, GPU, framework and cloud to the setups that fit your workload, then deploys the one you pick. Based in London. Contact: team@clearcompute.cc. - How it works: describe the workload (model, traffic, prompt and response lengths, latency target, region). ClearCompute checks every combination of model, GPU, inference engine, precision and cloud against what's available right now, ranks the setups that meet the target, and deploys the one the customer picks behind an OpenAI-compatible endpoint. - The customer decides: ClearCompute recommends and the customer chooses the setup and the cloud before anything is deployed. When a faster or cheaper setup appears, ClearCompute says so, and the customer decides whether to move. - Neutral: ClearCompute owns no GPUs and favours no cloud. It compares hyperscalers (AWS, Google Cloud, Azure) with specialist GPU clouds (such as RunPod, Lambda, Nebius, Hyperstack, Verda and CoreWeave). - Who it is for: businesses running open-source or fine-tuned models, and teams exploring whether self-hosting fits their workload. ## Pages - [Homepage](https://clearcompute.cc/index.md): what ClearCompute does and how it works - [Blog](https://clearcompute.cc/blog/index.md): all posts ## Blog posts - [What an H100 costs per hour in October 2026, and what the price leaves out](https://clearcompute.cc/blog/h100-cloud-prices.md): On 5 October 2026 one H100 listed at $2.69 to $12.29 per GPU-hour across nine clouds, with a $1.80 marketplace median. Where the spread comes from, and the costs the hourly rate leaves out. Published 2026-10-05, updated 2026-10-26. - [How much GPU memory does a 70B model need? Fitting it is the easy part](https://clearcompute.cc/blog/gpu-memory-70b-model.md): Llama 3.3 70B needs 141 GB of weights at BF16 and 73 GB at FP8, and every token of a user's context adds 320 KiB of KV cache. The arithmetic for how many users fit on an H100, two H100s, an H200 or a B200. Published 2026-10-05, updated 2026-10-26. - [Moving off a closed API: when do your own GPUs pay off?](https://clearcompute.cc/blog/self-hosting-vs-api-cost.md): We priced the same Llama 3.3 70B traffic three ways on 5 October 2026: a closed API, a per-token open-model API and dedicated H100s. Dedicated GPUs beat Sonnet-class prices above about 4% utilisation; for small or spiky traffic on a widely hosted model, the cheapest per-token hosts still cost less. Published 2026-10-05, updated 2026-10-26. ## Optional - [Full text of the site](https://clearcompute.cc/llms-full.txt)