anthropic/claude-opus-4.5 · Model identity and integrity report

Name: anthropic/claude-opus-4.5 relay integrity probe aggregate
Creator: BazaarLink

Last updated：Jul 30, 2026, 12:12 PM · Data window: past 90 days

Verification capability — fingerprint reference available

Family: anthropic
V3 sub-model reference: Available ✓
Checks: 知識截止 · 能力探針 · 拒絕模板 · 拒絕溫度參數
Family-fingerprint coverage: 60 weighted rules · 8 feature categories

Concrete reference answers, refusal templates, and individual signal weights are not published because disclosure would help spoofing endpoints evade detection.

Measured results (90 days)

36 runs · 14 hosts · 12 anomalies

Family-fingerprint consistency

  anthropic       33  ███████████████████████░░   92% (expected)
  openai           2  ██░░░░░░░░░░░░░░░░░░░░░░░    6% (suspicious)
  deepseek         1  █░░░░░░░░░░░░░░░░░░░░░░░░    3% (suspicious)

Observed V3 sub-model mapping

  → anthropic/claude-opus-4.5               23 (64%) ✓ correct sim=0.88
  → abstain                                  8 (22%) — low confidence
  → anthropic/claude-opus-4.6                2 ( 6%) ⚠ mismatch sim=0.79
  → anthropic/claude-sonnet-5                1 ( 3%) ⚠ mismatch sim=0.76
  → openai/gpt-5.4                           1 ( 3%) ⚠ mismatch sim=1.00
  → anthropic/claude-opus-4.7                1 ( 3%) ⚠ mismatch sim=0.85

Capability score distribution (auxiliary)

  0-49     ██████░░░░░░░░░░░░░░░░░░░    8
  50-69    ███░░░░░░░░░░░░░░░░░░░░░░    4
  70-89    █████████████████░░░░░░░░   24
  90-100   ░░░░░░░░░░░░░░░░░░░░░░░░░    0

Host integrity ranking

Host	Runs	Average score	Family match	Most common V3 match
https://ai.furry.vg/v1	6	43	100%	anthropic/claude-opus-4.5 (4)
https://turbo-api.com/v1	5	86	100%	anthropic/claude-opus-4.5 (5)
https://www.moyu.info/v1	5	80	100%	anthropic/claude-opus-4.5 (4)
https://ai-api.burgeon.cn/v1	4	48	100%	anthropic/claude-opus-4.5 (3)
https://api.gpt2share.com/v1	3	86	100%	anthropic/claude-opus-4.5 (3)
https://sarang.id/v1	3	45	67%	anthropic/claude-sonnet-5 (1)
https://api.qiyiguoai.com/v1	2	87	100%	anthropic/claude-opus-4.6 (2)
https://yunwu.ai/v1	2	82	100%	anthropic/claude-opus-4.5 (2)
https://api.sarang.id/v1	1	0	100%	—
https://api.quotaflow.ai/v1	1	85	100%	anthropic/claude-opus-4.5 (1)
https://api.lmuai.com/v1	1	0	100%	—
https://ai2.hhhl.cc/v1	1	0	0%	—
https://new-api.abrdns.com/v1	1	80	100%	anthropic/claude-opus-4.5 (1)
https://www.aitokens.link/v1	1	77	0%	openai/gpt-5.4 (1)

Recent anomalies

https://api.qiyiguoai.com/v1Jul 13, 2026, 4:45 PM

Self-claimed model: claude-opus-4-5-20251101
V2 predicted family: anthropic
V3 best match: anthropic/claude-opus-4.6 (sim 0.79)
Verdict: Sub-model downgrade

https://api.qiyiguoai.com/v1Jul 13, 2026, 4:17 PM

Self-claimed model: claude-opus-4-5-20251101
V2 predicted family: anthropic
V3 best match: anthropic/claude-opus-4.6 (sim 0.79)
Verdict: Sub-model downgrade

https://sarang.id/v1Jul 13, 2026, 12:07 AM

Self-claimed model: claude-opus-4.5
V2 predicted family: anthropic
V3 best match: anthropic/claude-sonnet-5 (sim 0.76)
Verdict: Sub-model downgrade

https://sarang.id/v1Jul 10, 2026, 10:58 AM

Self-claimed model: claude-opus-4.5
V2 predicted family: anthropic
V3 best match: —
Verdict: V3 low confidence · Abstain

https://sarang.id/v1Jul 10, 2026, 9:57 AM

Self-claimed model: claude-opus-4.5
V2 predicted family: deepseek
V3 best match: —
Verdict: V3 low confidence · Abstain

https://api.sarang.id/v1Jul 10, 2026, 9:53 AM

Self-claimed model: claude-opus-4.5
V2 predicted family: anthropic
V3 best match: —
Verdict: V3 low confidence · Abstain

https://ai.furry.vg/v1Jul 8, 2026, 4:09 AM

Self-claimed model: claude-opus-4-5-20251101
V2 predicted family: anthropic
V3 best match: —
Verdict: V3 low confidence · Abstain

https://ai.furry.vg/v1Jul 5, 2026, 4:35 PM

Self-claimed model: claude-opus-4-5-20251101
V2 predicted family: anthropic
V3 best match: —
Verdict: V3 low confidence · Abstain

https://api.lmuai.com/v1Jun 23, 2026, 2:15 PM

Self-claimed model: claude-opus-4-5
V2 predicted family: anthropic
V3 best match: —
Verdict: V3 low confidence · Abstain

https://ai2.hhhl.cc/v1Jun 5, 2026, 2:17 PM

Self-claimed model: claude-opus-4-5
V2 predicted family: openai
V3 best match: —
Verdict: V3 low confidence · Abstain

https://www.aitokens.link/v1May 26, 2026, 10:54 PM

Self-claimed model: claude-opus-4-5-20251101
V2 predicted family: openai
V3 best match: openai/gpt-5.4 (sim 1.00)
Verdict: Full swap

https://www.moyu.info/v1May 26, 2026, 4:21 PM

Self-claimed model: claude-opus-4-5-20251101
V2 predicted family: anthropic
V3 best match: anthropic/claude-opus-4.7 (sim 0.85)
Verdict: Sub-model downgrade

Methodology

BazaarLink Probe combines a Bayesian V2 family classifier across nine fingerprint categories with V3 sub-model references to verify endpoints claiming anthropic/claude-opus-4.5. Low family consistency, a different V3 match, or divergence from selfClaim is flagged as anomalous. BazaarLink Probe home.

Test an endpoint yourself →