Expired50% confidenceBenchmarkExact time
In comprehensive benchmark tests, Sol scores 55 (approximately 78.6%), Terra scores 44 (approximately 62.9%), and Luna scores 31 (approximately 44.3%).
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/10/2026
First Seen
Valid until: 10/8/2026(expired)
Sources
GPT-5.6 Three-Model Showdown: A Complete Breakdown of So, Terra, and Luna
bilibili宙流AI小工坊主7/9/2026
Related Claims
UnverifiedIn the proprietary benchmark PinBash 3 test, SOL scored approximately 78.6%, TERRA approximately 62.9%, and LUNA approximately 44.3%.76% similarPartially VerifiedIn the KingBench 3 evaluation, Soul scored 55 (approximately 78.6%), Tara scored 44 (approximately 62.9%), and Luna scored 31 (approximately 44.3%).71% similarUnverifiedClaude Opus 4.7 at an intermediate checkpoint achieves a pass rate of 57%, yet its final accuracy is only 39.8%71% similarUnverifiedIn the Tieba mini-site project evaluation, Sol ranked first with 83.95 points, GLM 5.2 ranked second with 76.59 points, Terra ranked third with 75.97 points, and Luna ranked fourth with 68.56 points.66% similarUnverifiedIn guided mode, the average end-to-end accuracy increased from 28.6% to 33.7%, and the ambiguity detection F1 score increased from 45.3% to 64.9%65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/471364API
curl https://kongchang.com/api/v1/knowledge/claims/471364MCP
get_claim(id=471364)