RAG Cost & Self-Hosting Calculator
Size the metered cost of a retrieval system — embeddings plus per-query tokens — and see how it climbs with usage, next to RAGSuite’s transparent per-seat licence and €0-per-token local models.
Metered cloud cost
At the rates and volume above. Provider list prices in USD.
The predictable alternative
RAGSuite is a transparent per-seat licence billed annually — and with local models via Ollama, the per-token meter stops entirely.
An estimate, not a quote. Real cost depends on your models, caching, reranking, retries and traffic shape. Cloud figures use the USD rates you set; RAGSuite prices are billed in EUR and shown unconverted — the comparison is about predictability, not an exchange rate. RAGSuite's Enterprise Edition is per-seat, billed annually — €300/user/year, lower per seat at volume (€240 at 50+ seats, €180 at 200+). Community is free with unlimited users. Enterprise services (white-label, SLA, air-gapped) are a separate agreement — talk to us.
Good to know
How does it estimate cost?
It converts your corpus to tokens (~1.33 tokens per word), prices a one-time embedding pass plus an ongoing re-embed for the share that changes, and adds per-query input (retrieved context) and output tokens at the rates you set. All arithmetic — no model calls.
Why are cloud figures in USD and RAGSuite in EUR?
LLM providers publish list prices in USD; RAGSuite is billed in EUR. We deliberately don’t convert currencies — the point isn’t an exchange-rate head-to-head, it’s the shape: metered cost climbs with usage, a self-hosted instance is a flat annual fee, and local models via Ollama cost €0 per token.
Are the prices accurate?
The rate fields are pre-filled with example list prices (June 2026) and are fully editable — set them to your provider’s current numbers, or paste your own quote. Nothing is fetched or stored; the defaults are a starting point, not a quote.
Want this modelled for your real workload?
Send yourself these figures and we’ll map them to a self-hosted RAGSuite deployment — transparent per-seat pricing billed annually, local models at €0 per token.
Thanks — we’ll be in touch within one business day (DE/EN).
Predictable per-seat pricing. No per-token meter.
RAGSuite self-hosts on your own infrastructure — transparent per-seat pricing billed annually, every LLM provider supported including local models via Ollama, and no metering surprises.