Llama 3.1 8B Instruct
Meta
meta-llama/llama-3.1-8b-instructLlama 3.1 8B Instruct is Meta's small open-weight instruction-tuned model, designed for fast, low-cost chat and general text tasks with a 128K context window. It suits lightweight applications and resource-constrained self-hosted deployments.
Cost rate
0.1x
Context
131K
Released
Jul 23, 2024
Input
Text
Output
Text
Support
ReasoningTool callingStructured outputs
Best at
The categories where this model ranks highest.
Medicine & Healthcare
#303 · top 84%
Legal & Government
#307 · top 84%
Mathematical
#313 · top 84%
Software & IT Services
#314 · top 80%
Performance
Median latency and throughput measured across recent requests.
Throughput137 tok/s
Latency