LLM API Cost Calculator
Same monthly budget, more usable tokens. See how much further your spend goes on Synthorai, with no platform fee, no payment fee, and prompt caching that stretches every dollar. Live prices from /api/pricing.
Your usage
⚡ The caching lever: set these to your workload
Cached context is billed at ~10% of input price, so the bigger the gap, the more you save.
Same budget · what it actually buys
| Platform | Fees | Buys tokens for | Tokens / mo |
|---|
“Buys tokens for” = how much of your budget reaches inference after fees; the rest is lost. OpenRouter uses your current cache rate, Synthorai your achievable rate. Top up $50, 10% off first month.
Estimates for comparison only. OpenRouter uses your current cache rate; Synthorai uses your achievable rate. Inference is priced at provider list price on both sides; the gap is OpenRouter's fees plus the caching you unlock. Actual usage depends on real traffic, model mix and cache behaviour.
Frequently asked questions
How much does OpenRouter charge in fees?
OpenRouter adds a platform fee (about 5.5%) plus payment processing (about 2.5%) when you buy credits, and about 5.5% on BYOK usage. Synthorai charges no platform fee and no payment-processing fee.
Is Synthorai cheaper than OpenRouter?
At the same provider list price, yes. Synthorai has no platform or payment fees, and prompt caching bills cache hits at about 10% of the input price, so the same monthly budget buys more tokens.
What is prompt caching and how does it lower cost?
Stable system prompts and retrieved context can be cached. Cache reads bill at roughly 10% of the normal input price, so a higher cache hit rate directly lowers your effective cost per request.
Does Synthorai support bring-your-own-key (BYOK)?
Yes. You can use your own provider keys with a 0% gateway surcharge, while OpenRouter charges about 5.5% on BYOK usage.