
CodeLlama 7B Instruct
Instruction-tuned 7B code model built on Llama 2, trained on 500B code tokens.
About
CodeLlama 7B Instruct is Meta's instruction-following code generation model, released on August 24, 2023 as part of the Code Llama family. It starts from Llama 2 7B, adds 500 billion tokens of code-specialized training, then instruction fine-tuning so it responds to natural-language requests about code, including generation, explanation, and review.
It was among the first widely adopted open-weight coding models and established the template for the code-model wave that followed. The 7B size runs on a single GPU, and the Instruct variant is the one Meta recommended for conversational coding assistance. Weights carry the Llama 2 Community License, a custom license with additional terms, so review before commercial use.
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 130; Trending score 82/100 from researcher assessment.