GPT-NeoX-20B
EleutherAI's 20B open model trained on the Pile, a landmark open LLM release.
About
GPT-NeoX-20B is EleutherAI's 20 billion parameter autoregressive language model, announced in February 2022 with weights released under Apache 2.0. Trained on the Pile with compute from CoreWeave, it was the largest dense autoregressive model with public weights at the time.
It became a foundational research artifact: a strong few-shot reasoner that powered countless open-source experiments before the chat-model era. EleutherAI notes it is a research artifact, not hardened for production deployment. Development was led by Sid Black, Stella Biderman, and Eric Hallahan.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 110; Trending score 80/100 from researcher assessment.