Pricing°
Llama 3.3 70B Instruct cuts input and output pricing
Llama 3.3 70B Instruct pricing on OpenRouter has dropped. Input moves from $0.22 to $0.1 per 1M tokens, and output from $0.5 to $0.32 per 1M tokens.
Pricing5 Oct 2026Meta
What it means for a SaaS brand
Cheaper open-weight inference makes it more affordable to run high-volume AI search visibility monitoring, so widen your prompt sets and run them more often rather than sampling a handful of queries. If you use Llama 3.3 70B Instruct in your own product, recheck your margin assumptions and unit costs this week.
Source
Every entry here names where it came from, and we publish nothing we have not read at the source.
Read the originalChanges like this decide who gets cited
We measure how often AI answers name your brand, track what moves it, and tell you what to do next.