Coverage (30d)
1vs2
This Week
1vs2
Evidence
1 articlesRelationships
1Timeline
Qwen3-30B-A3B2026-04-16
Achieved 73.4% on SWE-bench Verified, beating predecessor and competing with Claude 3.5 Sonnet on vision tasks
Qwen3-30B-A3B2026-03-16
Achieved 79.6% invariant responses in semantic invariance testing, outperforming larger models up to 405B parameters.
Qwen 3.5 Medium2026-02-25
Outperformed its 235B parameter predecessor while using 7x fewer active parameters per token
Qwen 3.5 Medium2026-02-24
Demonstrated remarkable efficiency gains through architectural improvements
Qwen 3.5 Medium2026-02-01
Recently released model used for performance comparison
Ecosystem
Qwen3-30B-A3B
usesSemantic Invariance1 src
competes withClaude 3.5 Sonnet1 src
competes withQwen 3.5 Medium1 src
Qwen 3.5 Medium
competes withQwen2.5-235B1 src