Meta
Efficient

CodeLlama 13B Instruct

Code

Smaller CodeLlama. Mostly superseded by Qwen2.5-Coder 14B and Phi-4 for new self-hosted code work.

CodeLlama 13B Instruct is a efficient AI model from Meta. It costs $0.2200 per million input tokens and $0.2200 per million output tokens (blended $0.2200/M), with a 16,000-token context window.

RatesUSD per million tokens. Blend is 70% input, 30% output, our calculation.

INPUT
$0.2200/M
per million input tokens
OUTPUT
$0.2200/M
per million output tokens
BLENDED 70/30
$0.2200/M
unchanged since 3 May
CONTEXT
16,000
tokens

Hand-maintained from Meta's published pricing; no check date on record.

What it is good at
  • Open weights
  • Single-GPU friendly
Typical use cases
  • Legacy CodeLlama deployments

Benchmarks

vs. best public score
HumanEval42.7%
Python function synthesis from docstrings. Source: Meta (self-reported), HumanEval, read 2026-11-10.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.

More from Meta

See all 20 →

Frequently asked questions

How much does CodeLlama 13B Instruct cost?

CodeLlama 13B Instruct costs $0.2200 per million input tokens and $0.2200 per million output tokens, for a blended reference rate of $0.2200 per million tokens.

What is CodeLlama 13B Instruct's context window?

CodeLlama 13B Instruct supports up to 16,000 tokens of context in a single request.

What is CodeLlama 13B Instruct best for?

CodeLlama 13B Instruct is well suited to Open weights, Single-GPU friendly and Legacy CodeLlama deployments.

Who makes CodeLlama 13B Instruct?

CodeLlama 13B Instruct is developed and served by Meta. It was released in Aug 2023.

Terms used on this page