It is insanely more expensive on replica and they don't have the 70b model yet which will make it even more prohibitive.
As for pricing, that model's pages says: "Predictions run on Nvidia A100 (80GB) GPU hardware. Predictions typically complete within 17 seconds."
And the pricing page (https://replicate.com/pricing) says Nvidia A100 (80GB) GPU hardware costs $0.0032 per second.
So Llama 2 70B would "typically" cost under 17 x 0.0032 = $0.0544 per run.