Artificial Analysis thread (reply 3): knowledge ‘a relative weakness’ for Muse Wonder, with low hallucination rates
Artificial Analysis @ArtificialAnlysX
Why it matters
Independent knowledge and hallucination results for both Muse Wonder models.
Summary
Posted Oct 12, 2026 (17:25 UTC). Third reply in Artificial Analysis’ thread (root: x-ArtificialAnlys-2109697083519447118). ‘Knowledge remains a relative weakness for both models, but this is mitigated by very low hallucination rates.’ On AA-Omniscience, Muse Wonder Full (max) scores 30 and Muse Wonder (max) 25, below Claude Opus 5.5 (46), GPT-6 Astra (43), Gemini 4 Argon (42) and Claude Sonnet 5.5 (32). Wonder Full lowered its hallucination rate from 33% to 23%, the second lowest of any model scoring 45 or more on the Intelligence Index, behind Gemini 4 Argon (high, 15%); Muse Wonder (max) is third at 27%.
Metrics as of 2026-10-13 via api.fxtwitter.com: 5,134 views, 42 likes, 1 reposts.
Archived text
Knowledge remains a relative weakness for both models, but this is mitigated by very low hallucination rates. On AA-Omniscience, Muse Wonder Full (max) scores 30 and Muse Wonder (max) scores 25, level with Muse Spark 1.3 (max), below e.g. Claude Opus 5.5 (46), GPT-6 Astra (43), Gemini 4 Argon (42), and Claude Sonnet 5.5 (32)
Muse Wonder Full (max)’s 5-point gain comes from lowering its hallucination rate (33% to 23%), which places it among the least-hallucinating models tested: it has the second-lowest hallucination rate of any model scoring 45 or more on the Intelligence Index, behind only Gemini 4 Argon (high, 15%). Muse Wonder (max) is third at 27%
