Best local LLM for agents
Top 7 open-source picks for autonomous agents and tool use, ranked by benchmark performance and real-world fit. Updated monthly.
Qwen 3 30B-A3B
Alibaba's Qwen 3 MoE with 30B total and just 3B active parameters, supporting hybrid thinking mode. MMLU 81.4, AIME24 80.4, 100+ languages, Apache 2.0.
gpt-oss 20B
OpenAI's compact open-weight MoE with 3.6B active out of 21B total parameters. Matches o3-mini on a laptop-class GPU under Apache 2.0.
ERNIE 4.5 21B-A3B Thinking
Baidu's compact reasoning MoE with 3B active parameters out of 21B total. Fast inference thanks to the small active set, with Chinese-language strength.
Kanana 2 30B-A3B Thinking
Kakao's agentic 30B MoE (3B active) with native hybrid thinking and Korean-first training. Apache 2.0 with MLA attention and 131k context.
DeepSeek R1 Distill 32B
The 32B DeepSeek R1 distill — the best accessible open-weight reasoner we've tested. Explicit chain-of-thought, MIT-licensed, runs on a single 24GB GPU.
Qwen 3 32B
Alibaba's 32B dense flagship with thinking mode, scoring 65.5 on MMLU-Pro and 39.8 on SuperGPQA. The strongest general-purpose Qwen 3 dense model before stepping up to the MoE.
QwQ 32B
Alibaba's dedicated 32B reasoner, trained with reinforcement learning rather than distillation. Hits 79.5 on AIME24 and 90.6 on MATH-500 — a direct Apache-licensed alternative to DeepSeek R1.