▸ __BREADCRUMB__ usage
new: genuinely free AI models are live — try surp/free + see live budgets · token-gating prototype · vote on SRP

cheapest LLM API

price check: 150+ models, live.

Looking for the cheapest LLM API? Here's the thing: the cheapest model changes constantly. Sellers on the Surplus Intelligence marketplace reprice in real time, and new models launch all the time. A static comparison is stale the moment it's published.

surp.ivc.lol solves this differently: it's an LLM API that routes to whatever is cheapest at request time. You don't pick a provider — you pick a class of work ("coding", "chat", "reasoning") and the gateway picks the cheapest model serving that class right now. You pay the live market price plus a 5% gateway fee.

live pricing vs retail

examples from the live marketplace (per 1M tokens):

modelretail list pricevia surp.ivc.lolsavings
deepseek-v4-flash~$0.20~$0.04~80%
qwen3-coder-turbo~$1.00~$0.42~58%
gpt-5.6-terra~$1.75~$0.18~90%
glm-5.2~$0.60~$0.03~95%

these move every minute — see the live ticker, top-5 leaderboards, or browse all models.

why we're cheaper than retail

  • Marketplace pricing — sellers compete on price; you get the best offer, not a fixed list price.
  • Dynamic routing — no overpaying for a "premium" model when a cheaper one is just as good for your task.
  • No subscription overhead — no unused quota to subsidize, no billing minimums.
  • Open models included — DeepSeek, Qwen, GLM and other open-weight models cost a fraction of closed frontier models for most tasks.

how to get the cheapest LLM API right now

curl -X POST https://surp.ivc.lol/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"surp/chat","messages":[{"role":"user","content":"hello"}],"max_tokens":50}'

surp/chat = cheapest general chat model at this second. surp/best-coding = cheapest coder. Find the right model for your task »

related: what is x402? · x402 LLM API · pay-per-request LLM API · compare models · top 5 leaderboards