---
title: "Llama 3.1 8B Benchmarks - RankedAGI"
description: "Detailed benchmark and metadata record for Llama 3.1."
source: "https://rankedagi.com/models/llama-3.1-8b"
---

# Llama 3.1

| Field | Value |
| --- | --- |
| Organization | Meta |
| License | Open Source |
| Size | 8B |
| Version | Latest |
| Released | 2024-07-23 |
| Context window | 128K |
| Knowledge cutoff | 2023-12-01 |
| Input cost per million tokens | $0.3 |
| Output cost per million tokens | $0.6 |
| Last updated | 2025-04-16 22:14:46.146909+00 |

## Benchmarks

| Benchmark | Category | Value | Description | Source |
| --- | --- | --- | --- | --- |
| RankedAGI Coding | coding | 19.8% | RankedAGI Coding Score |  |
| RankedAGI Agentic | Agents | 19.0% | RankedAGI Agentic Score |  |
| ChatArena (LMSYS) | coding | 1174 | ChatArena (LMSYS) Coding ELO Score with Style Control |  |
| LiveBench Coding Score (old) | coding | 19.7% | LiveBench Coding Score old version |  |
| RankedAGI Reasoning | reasoning | 28.0% | RankedAGI Reasoning Score |  |
| GPQA Diamond | reasoning | 27.0% | Generalized Prefix Question Answering Score (Reasoning) PhD Level Reasoning |  |
| Text Arena | general | 1171 | ChatArena (LMSYS) ELO Score | https://arena.ai/leaderboard/text |
| HumanEval | coding | 72.6% | HumanEval coding benchmark |  |
| HumanEval+ | coding | 62.8% | HumanEval+ benchmark |  |
| MMLU Pro | knowledge | 44.3% | Graduate-level visual problem-solving |  |
| MMLU | knowledge | 71.0% | Image Reasoning |  |
| Hughes Hallucination Evaluation | safety | 5.4% | Computed using Vectara's Hallucination Evaluation Model, also known as HHEM. This evaluates how often an LLM introduces hallucinations when summarizing a document. | https://github.com/vectara/hallucination-leaderboard |
| Artificial Analysis Quality Index | general | 53 | Artificial Analysis Quality Index (third-party composite, deprecated) |  |
| RankedAGI Math | math | 37.0% | RankedAGI Math Score |  |
| RankedAGI Overall | general | 31.7% | Overall RankedAGI score |  |