Meta · Meta AI (Llama)
Llama 3 70B
SupersededApril 2024 quality jump; 8K context (later extended to 128K in 3.1). Introduced GQA for faster inference. Superseded by 3.1/3.3.
About this version
Llama 3 70B is a 70‑billion‑parameter language model from Meta released in April 2024 as part of the Llama 3 family. The version expanded the context window to 8 K tokens, added grouped query attention for faster inference, and was noted for a quality improvement over earlier releases; it has since been superseded by later variants such as Llama 3.1 and Llama 3.3.
- Context window
- 8K ctx
- Tier
- Legacy
- Released
- Apr 1, 2024
Release & change timeline
1 event- Release
Llama 3 released
8B and 70B; major quality jump, 8K context (extended to 128K in 3.1). Month known, exact day not stated.
Report an issue with Meta AI (Llama)
Posts to your status feed
Pick the closest match below, edit the body, and post. Your report carries the #meta tag automatically so it surfaces here + in the trending-tags rail. Reporting also follows Meta AI (Llama) so you’ll get status updates.
Part of the Meta AI (Llama) version timeline. Catch Me Up — everything that changed across AI & smart home ↗