Price comparisons
One page per question, generated from the data. Every number carries its verification time and links to the provider's own page. List prices, not transacted prices.
GPU price per hour
- H100 price per hour: USD 4.41 median, per GPU-hour · 2026-10-23 (UTC) · n=17
- H200 price per hour: USD 5.89 median, per GPU-hour · 2026-10-23 (UTC) · n=13
- B200 price per hour: USD 8.36 median, per GPU-hour · 2026-10-23 (UTC) · n=11
- A100 80GB price per hour: USD 2.13 median, per GPU-hour · 2026-10-23 (UTC) · n=7
- L40S price per hour: USD 1.61 median, per GPU-hour · 2026-10-23 (UTC) · n=10
- RTX 4090 price per hour: not computed median, thin coverage: 1 verified providers with an on-demand RTX 4090 list price (minimum 4)
Cheapest VPS by size
- Cheapest 1 vCPU / 1 GB VPS in the EU: EUR 0.438 per month · 2026-10-23 (UTC) · n=7
- Cheapest 2 vCPU / 4 GB VPS in the EU: EUR 4.49 per month · 2026-10-23 (UTC) · n=7
- Cheapest 4 vCPU / 8 GB VPS in the EU: EUR 6.49 per month · 2026-10-23 (UTC) · n=6
- Cheapest 8 vCPU / 16 GB VPS in the EU: EUR 19.99 per month · 2026-10-23 (UTC) · n=6
- Cheapest 1 vCPU / 1 GB VPS in the US: not computed thin coverage: 1 verified providers with a qualifying US plan (minimum 3)
- Cheapest 2 vCPU / 4 GB VPS in the US: not computed thin coverage: 1 verified providers with a qualifying US plan (minimum 3)
- Cheapest 4 vCPU / 8 GB VPS in the US: not computed thin coverage: 1 verified providers with a qualifying US plan (minimum 3)
- Cheapest 8 vCPU / 16 GB VPS in the US: not computed thin coverage: 1 verified providers with a qualifying US plan (minimum 3)
Open models across inference hosts
- Llama 3.3 70B Instruct API price across hosts: USD 0.10 cheapest host, per 1M input tokens · 2026-10-23 (UTC) · n=10
- DeepSeek V3.2 API price across hosts: USD 0.2088 cheapest host, per 1M input tokens · 2026-10-23 (UTC) · n=11
- Qwen3 235B A22B Instruct 2507 API price across hosts: USD 0.09 cheapest host, per 1M input tokens · 2026-10-23 (UTC) · n=3
- gpt-oss-120b API price across hosts: USD 0.03 cheapest host, per 1M input tokens · 2026-10-23 (UTC) · n=21
LLM API price per model (first-party)
- Claude Fable 5.1 API price
- Claude Haiku 4.5 API price
- Claude Haiku 5.5 API price
- Claude Opus 4.5 API price
- Claude Opus 5.5 API price
- Claude Sonnet 4.5 API price
- Claude Sonnet 5.5 API price
- DeepSeek Flash API price
- DeepSeek V4 Pro API price
- Gemini 2.5 Flash API price
- Gemini 2.5 Flash Lite API price
- Gemini 2.5 Pro API price
- Gemini 3.1 Flash Lite API price
- Gemini 3.5 Flash Lite API price
- Gemini 3.6 Flash API price
- Gemini 3.8 Flash API price
- Gemini 4 Argon API price
- Mistral Large API price
- GPT-6 Astra API price
- GPT-6 Luna API price
- GPT-6.1 Sol API price
- Grok 4.7 API price