Qwen 3 Coder 30B A3B Instruct — benchmark results

Alibaba's open 30B A3B MoE coder (3.3B active) tuned for agentic coding, with 256K native context extendable to 1M (July 2025). Provider: Alibaba. Released 2025-07-23. Access: Open.

Unified ELO 1486 ± 29, rank #857 of 1776 rated models, from 51 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA MATH-50089.33Accuracy (%)64.4
AA Terminal-Bench Hard15.15Accuracy (%)52.3
SnakeBench22.7TrueSkill Rating52.2
AA-LCR29Score (self-reported)48.9
AA LiveCodeBench40.32Pass@1 (%)48
AA Omniscience - Software Engineering (SWE) - Julia12Accuracy (%)47.9
ScratchWorld - Debug30Success Rate (%)46.2
ScratchWorld - Overall9.64Success Rate (%)46.2
Artificial Analysis Intelligence Index13.61Intelligence Index45.4
UGI - Writing31.8Writing Score43.9
AA TAU-2 Bench34.5Accuracy (%)42.4
ScratchWorld - Create10Success Rate (%)42.3

Interactive version: theaggregate.ai/model?slug=qwen-3-coder-30b-a3b-instruct · How the rankings work · Data refreshed daily, snapshot 2026-07-22.