Meta: Llama 3 8B Instruct

Meta: Llama 3 8B Instruct

meta-llama · Released Apr 18, 2024
35
Our Score

Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases. It has demonstrated strong performance compared to leading closed-source models in human evaluations. Usage of this model is subject to Meta's Acceptable Use Policy.

$0.03 / 1M Input Price
$0.04 / 1M Output Price
8,192 tokens Context Window
16,384 tokens Max Output
8B Parameters

Capabilities

Tool Use Function Calling

Architecture

ModalityText → Text
TokenizerLlama3
Instruct Typellama3
Parameters8B

Performance Indices

Source: Artificial Analysis

6.4 Intelligence Index
4 Coding Index

Benchmark Scores

Evaluations

GPQA Diamond 29.6%
Graduate-level scientific reasoning
HLE 5.1%
Humanity's Last Exam
MMLU Pro 40.5%
Multi-task language understanding
LiveCodeBench 9.6%
Live coding evaluation
SciCode 11.9%
Scientific computing
MATH 500 49.9%
Mathematical problem-solving
IFBench 24.6%
Instruction following

Benchmark data from Artificial Analysis and Hugging Face

Model Information

OpenRouter ID meta-llama/llama-3-8b-instruct
Providermeta-llama
Model FamilyLlama 3
Release Date April 18, 2024
Context Length8,192 tokens
Max Completion16,384 tokens
Status Active

Pricing

Token Type Cost per 1M tokens Cost per 1K tokens
Input $0.03 $0.000030
Output $0.04 $0.000040

Live Performance

Live endpoint metrics — refreshed every 30 minutes.

99.1%
Avg Uptime
228ms
Best Latency (TTFT)
46 tok/s
Best Throughput
3/4
Active Endpoints
Available via: DeepInfra, Novita, Together, Cloudflare