Unverified50% confidenceBenchmarkExact time
NuExtract3 scored 0.651 in NuMind's internal test of about 600 documents, versus 0.417 for the base model Qwen3.5-4B; on ExtractBench it averaged 82.35, but neither states coverage of financial report documents.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/13/2026
First Seen
Valid until: 1/11/2027
Sources
Related Entities
Related Claims
UnverifiedQwen3 27B在Kanban看板应用测试中消耗近87%上下文，一次生成出无bug的成果，支持拖拽、排序、筛选、归档等功能64% similarUnverifiedIn the Terminal Bench test, Qwen 3.6 27B scored approximately 60.7, while Muse Glimmer scored only 51.7, falling significantly behind.63% similarUnverifiedOn Bespoke Labs' leakage-level evaluation, Nimble reaches 90.12, while its base model Qwen 3.5 9B scores only 66.36.63% similarUnverifiedQwen3.8-27B scored 91.1 on OmniDocBench v1.5, the strongest document parsing score among the generative models in this evaluation.62% similarUnverifiedMuse Glimmer scores 75.5 on the MCP Atlas tool-calling benchmark, leading Qwen 3.6; scores 74.6 on agentic search tasks; and scores 94.7 on the AIME mathematical reasoning benchmark.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1013929API
curl https://kongchang.com/api/v1/knowledge/claims/1013929MCP
get_claim(id=1013929)