Model comparisons

Model vs model, from the live catalog.

Each page puts two models or two families side by side: per-1M prices, cached-input rates, context windows, maximum output and capability flags, read from the RouterPlex catalog on every request. Worked costs show the arithmetic. Benchmark rows appear only where a cited source exists. Every model here runs on one OpenAI-compatible key.

Prices as of 17 November 2026

One key

Switch models by changing one string.

Top up from $5 with no subscription. Looking at gateways instead of models? See RouterPlex vs other gateways.

AI Model Comparisons: API Price and Context · RouterPlex