November 26, 2025
Cheapskate or cheapscam?
API that auto-routes to the cheapest AI provider (OpenAI/Anthropic/Gemini)
Cheapest AI router promises big savings — commenters shout “typo!” and “OpenRouter at home”
TLDR: A new API promises to auto-route AI requests to the cheapest provider for big savings, but commenters erupted over confusing pricing units, “OpenRouter clone” vibes, and sketchy site details. Devs want lower costs, yet the community is demanding clear numbers and real security before trusting it.
An eyebrow-raising new tool claims it can auto-route your AI requests to the cheapest provider — OpenAI, Anthropic, or Google Gemini — with up to 90% savings, one simple endpoint, and automatic fallbacks. The founder says they built it “out of frustration” with messy pricing, promising real-time price checks, 30 free requests, and Stripe billing. But the crowd isn’t quietly cheering — they’re roasting.
The strongest reaction: confusion and panic over the posted prices. One user saw “$0.50 per 1,000 input tokens” and joked they’d be bankrupt in an hour, while another lined it up against “GPT-5.1” pricing and asked bluntly: is this a typo? The meme brigade showed up with “Mom, I want OpenRouter. We have OpenRouter at home,” accusing this service of being a knockoff of existing model routers.
Then came the trust drama: commenters say the landing page looks “vibe coded,” with outdated model names, a dead security link, and even a visible dashboard without logging in. Skeptics argue that the biggest, priciest queries need premium models — the exact ones this router claims to avoid — so savings might be smoke and mirrors. Devs want cheaper AI, yes, but they want clarity, accuracy, and security even more.
Key Points
- •Single API auto-routes AI requests to the cheapest provider among OpenAI, Anthropic, and Google Gemini.
- •Claims cost savings of up to 90–99% versus using premium models directly.
- •Provides automatic fallback to keep apps online if a provider fails.
- •Offers unified responses, real-time pricing monitoring, and visibility into provider used, tokens, and costs.
- •Billing is per request via Stripe, with 30 free requests and no card required; usage visible in a customer dashboard.