Llama 3.2 3B Instruct

by Meta

Llama 3.2 3B Instruct is Meta's small but capable text model, released September 2024 for efficient inference at competitive quality. With 3 billion parameters and a 32,768-token context window, it sits above the 1B model and suits tasks where base generation quality matters more than minimal footprint. It handles general-purpose text understanding, summarization, simple coding tasks, and dialogue, and runs efficiently on consumer hardware, making it accessible for researchers, startups, and cost-sensitive deployments. It is useful for question-answering systems, content moderation, text classification, and customer support automation at lower operational cost than larger models. The compact context window suits shorter documents and single-turn interactions, and it remains a popular balance between capability and efficiency.

Key info

Input
Output
Features
Context window
33K
Max output
32K

Benchmarks

Independent benchmark scores — composite indices for reasoning, coding, and math, plus individual eval scores where available.

Global rank#522 of 596 LLMs
TierEfficient
Output speed0 tok/s
First token0.00s
Intelligence Index3.9
Math Index3.3
Reasoning & knowledge
MMLU-Pro
35%
GPQA Diamond
26%
Humanity's Last Exam
5%
Long-context reasoning
3%
Coding
LiveCodeBench
8%
SciCode
5%
Agentic & tool use
τ²-Bench Telecom
21%
Math & instruction following
AIME 2025
3%
IFBench
26%

Available routes

No routes currently available — Llama 3.2 3B Instruct isn't routed through the Opper gateway right now. It may return.

Contact us about this model →

Available models from Meta

Start building with 700+ models

One API key. Every major provider. Up and running in minutes.

Get startedView Documentation
Llama 3.2 3B Instruct by Meta — not currently on Opper | Opper AI