Unverified50% confidenceBenchmarkExact time
The Qwen4 preview version scored only 35.9 on the HLE reasoning leaderboard, losing to a competitor's enhanced version that scored 40.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/7/2026
First Seen
Valid until: 1/5/2027
Sources
Related Entities
Related Claims
UnverifiedOn pure reasoning tasks, Qwen3 27B scores only 30.8 on HLE, notably behind Opus's score of 40.0.74% similarUnverifiedIn the Terminal Bench test, Qwen 3.6 27B scored approximately 60.7, while Muse Glimmer scored only 51.7, falling significantly behind.68% similarUnverifiedQwen3-1.7B's overall causal faithfulness is only 54.8%, far below that of Qwen3-4B.68% similarUnverifiedThe IQ4XS quantized version of Qwen3 27B shows only approximately 6% performance degradation compared to the original full version.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/985982API
curl https://kongchang.com/api/v1/knowledge/claims/985982MCP
get_claim(id=985982)