CodeLlama 70B

Name: CodeLlama 70B
Rating: 60.7 (25000 reviews)
Author: Meta

#168CodeOpen Weights

Meta

Code Llama 70B is a code-specialized large language model built on Llama 2. It supports code generation, completion, and infilling across popular programming languages.

Running this yourself: likely needs a high-memory cloud gpu.

Model updates refreshed3h agoMay 21, 2026news + changelog

Website View Updates Get API Access

60.7

Quality Score

1155

Arena ELO

70B

Parameters

100K

Context

Benchmarks and Competitive Signal

Structured

Use this section to answer one simple question first: how much outside evidence do we have that this model performs well? Structured benchmark scores appear first, then official provider evidence, then live arena signal.

This model has normalized benchmark rows, so scores here are directly comparable across benchmark sources.

SWE-Benchcoding

42.4