Unverified50% confidenceFactExact time
The thinking intensity levels set for the test were: Max for Sol and DeepSeek, and High for Luna.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/29/2026
First Seen
Valid until: 12/28/2026
Sources
Related Entities
Related Claims
UnverifiedIn the agent ultimate exam covering long-horizon workflows across 55 professional domains, Sol scored 53.6, surpassing the benchmarked Fable 5 by 13.1 points.67% similarUnverifiedIn Claude Code, selecting Low, Medium, or High effort levels all map to the same level of effectiveness; only selecting Max Effort or higher maps to the strongest capability.65% similarUnverified初代 DeepSeek Math 7B 在 MATH 竞赛基准测试中以70亿参数达到 51.7% 的正确率64% similarUnverifiedMAI Thinking-E performs strongly in reasoning-intensive domains such as mathematics and science, but in coding tests it only reached the level of DeepSeek V3.2.62% similarUnverifiedBenchmark results show that Sol and Sol Ultra outperform competitor Misos 5 in most areas, and Tara outperforms Fable 5.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/958965API
curl https://kongchang.com/api/v1/knowledge/claims/958965MCP
get_claim(id=958965)