Expired85% confidenceFactTime unknown
Qwen3.5 scored 76.5 on the IFBench instruction-following evaluation, surpassing GPT-5.2's 75.4 and Claude 4.5 Opus's 58.
1
Sources
85%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
Valid until: 8/30/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedQwen3.6 35B's score on SWE Bench Verified is only 1.6 percentage points behind Claude Opus 4.6.77% similarUnverifiedQwen3.5 scored 91.3 on the AIME 2026 mathematics competition evaluation, while GPT-5.2 scored 96.7.77% similarUnverifiedQwen3-4B is already on par with DeepSeek V3 or GPT-4o on multiple reasoning evaluation metrics.77% similarUnverifiedQwen3.5 scored 28.7 on the HLE doctoral-level comprehensive question evaluation, while GPT-5.2 scored 35.5.76% similarUnverifiedQwen 3's flagship version matches or surpasses GPT-4o and Claude 3.5 Sonnet in multiple benchmark tests.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/18080API
curl https://kongchang.com/api/v1/knowledge/claims/18080MCP
get_claim(id=18080)