← Back to model

cantus · Model identity and integrity report

Last updated：Sep 26, 2026, 11:45 AM · Data window: past 90 days

Measured results (90 days)

8 runs · 3 hosts · 8 anomalies

Family-fingerprint consistency

  qwen             5  ████████████████░░░░░░░░░   63% (suspicious)
  anthropic        3  ██████████░░░░░░░░░░░░░░░   38% (suspicious)

Observed V3 sub-model mapping

  → abstain                                  4 (50%) — low confidence
  → qwen/qwen3-max                           2 (25%) ⚠ mismatch sim=0.70
  → qwen/qwen3-next-80b-a3b-instruct         2 (25%) ⚠ mismatch sim=0.68

Capability score distribution (auxiliary)

  0-49     ██████████░░░░░░░░░░░░░░░    3
  50-69    ░░░░░░░░░░░░░░░░░░░░░░░░░    0
  70-89    ████████████████░░░░░░░░░    5
  90-100   ░░░░░░░░░░░░░░░░░░░░░░░░░    0

Host integrity ranking

Host	Runs	Average score	Family match	Most common V3 match
https://api.gnrt.dev/v1	4	56	0%	qwen/qwen3-max (2)
https://ai.kaiplay.my.id/v1	3	54	0%	—
https://ai.redromance.pro/v1	1	0	0%	—

Recent anomalies

https://api.gnrt.dev/v1Aug 25, 2026, 7:45 PM

Self-claimed model: qd/cantus
V2 predicted family: qwen
V3 best match: qwen/qwen3-max (sim 0.70)
Verdict: Full swap

https://api.gnrt.dev/v1Aug 25, 2026, 11:52 AM

Self-claimed model: qd/cantus
V2 predicted family: qwen
V3 best match: qwen/qwen3-max (sim 0.71)
Verdict: Full swap

https://ai.redromance.pro/v1Aug 23, 2026, 4:23 PM

Self-claimed model: Cantus
V2 predicted family: anthropic
V3 best match: —
Verdict: V3 low confidence · Abstain

https://ai.kaiplay.my.id/v1Aug 10, 2026, 1:24 PM

Self-claimed model: qd/cantus
V2 predicted family: qwen
V3 best match: —
Verdict: Family mismatch

https://ai.kaiplay.my.id/v1Aug 10, 2026, 6:51 AM

Self-claimed model: qd/cantus
V2 predicted family: anthropic
V3 best match: —
Verdict: V3 low confidence · Abstain

https://ai.kaiplay.my.id/v1Aug 9, 2026, 11:57 AM

Self-claimed model: qd/cantus
V2 predicted family: anthropic
V3 best match: —
Verdict: Family mismatch

https://api.gnrt.dev/v1Aug 4, 2026, 12:42 PM

Self-claimed model: qd/cantus
V2 predicted family: qwen
V3 best match: qwen/qwen3-next-80b-a3b-instruct (sim 0.65)
Verdict: Full swap

https://api.gnrt.dev/v1Jul 31, 2026, 3:26 AM

Self-claimed model: qd/cantus
V2 predicted family: qwen
V3 best match: qwen/qwen3-next-80b-a3b-instruct (sim 0.72)
Verdict: Full swap

Methodology

BazaarLink Probe combines a Bayesian V2 family classifier across nine fingerprint categories with V3 sub-model references to verify endpoints claiming cantus. Low family consistency, a different V3 match, or divergence from selfClaim is flagged as anomalous. BazaarLink Probe home.

Test an endpoint yourself →