Meta
Llama-3.2-3B-Instruct benchmark results
4 results on 4 benchmarks from 4 sources.
Also listed as: llama-3.2-3b-instruct · Llama-3.2-3B-Instruct (FC) · Llama-3.2-3B · meta-llama/Llama-3.2-3B-Instruct · meta-llama/Llama-3.2-3B
All benchmark results
| Benchmarks | Score | Rank | Setting | Source type | Date |
|---|---|---|---|---|---|
| LMArena Text | 1167 | 367 | Overall, style control on; 7936 votes; 95% CI 1159–1174“llama-3.2-3b-instruct” | Official leaderboard | Oct 8, 2026 |
| Open LLM Leaderboard v2 | 24.2% | 1768 | Average of 6 normalised benchmarks; bfloat16; 3.213B params“meta-llama/Llama-3.2-3B-Instruct” | Official leaderboard | Mar 13, 2025 |
| MMLU-Pro | 22.2% | 229 | Overall accuracy; data source: TIGER-Lab; 3B“Llama-3.2-3B” | Official leaderboard | Mar 11, 2026 |
| BFCL v4 (function calling) | 21.9% | 98 | Overall accuracy; native function calling (FC)“Llama-3.2-3B-Instruct (FC)” | Official leaderboard | Oct 17, 2026 |
FAQ
Where do these Llama-3.2-3B-Instruct scores come from?
From Berkeley Function Calling Leaderboard (BFCL), LMArena leaderboard dataset, MMLU-Pro leaderboard (TIGER-Lab), and Hugging Face Open LLM Leaderboard v2 (archived). Each row links to the original leaderboard, dataset or report, and vendor-reported numbers are labelled.