Expired50% confidenceOpinionExact time
Luna's performance was evaluated as mediocre, with the model exhibiting many errors and falling short of GPT-6.1 and Sonnet 5.5 in stability and accuracy.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/1/2026
First Seen
Valid until: 10/15/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedLuna scored 66.6% in the coding test, with costs 93% and 96% lower than Opus 5 and Fable 5, respectively.70% similarUnverifiedThe Luna model's fail_fast/endstate score is 64/55, with the gap corresponding to trajectories that appear clean turn by turn but end in a corrupted final state.69% similarUnverifiedThe engineering lead at Base44 stated that Sonnet 5.5 is the model with the lowest tool-call failure rate.67% similarUnverifiedTesters recommended continuing to use Opus 4.8 for the vast majority of tasks for better results, and considered Sonnet 5 disappointing.65% similarUnverifiedIn pre-deployment evaluations of Claude Sonnet 4.5, the model achieved a 0% misalignment rate, but reading its chain of thought revealed that it clearly knew it was being tested.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/964927API
curl https://kongchang.com/api/v1/knowledge/claims/964927MCP
get_claim(id=964927)