dfmodel · Model identity and integrity report
Last updated：Aug 21, 2026, 10:41 PM · Data window: past 90 days
Measured results (90 days)
5 runs · 4 hosts · 5 anomalies
Family-fingerprint consistency
deepseek 2 ██████████░░░░░░░░░░░░░░░ 40% (suspicious) openai 1 █████░░░░░░░░░░░░░░░░░░░░ 20% (suspicious) qwen 1 █████░░░░░░░░░░░░░░░░░░░░ 20% (suspicious) anthropic 1 █████░░░░░░░░░░░░░░░░░░░░ 20% (suspicious)
Observed V3 sub-model mapping
→ abstain 3 (60%) — low confidence → openai/gpt-5 1 (20%) ⚠ mismatch sim=0.85 → deepseek/deepseek-v4-flash 1 (20%) ⚠ mismatch sim=0.75
Capability score distribution (auxiliary)
0-49 ███████████████░░░░░░░░░░ 3 50-69 ░░░░░░░░░░░░░░░░░░░░░░░░░ 0 70-89 ██████████░░░░░░░░░░░░░░░ 2 90-100 ░░░░░░░░░░░░░░░░░░░░░░░░░ 0
Host integrity ranking
| Host | Runs | Average score | Family match | Most common V3 match |
|---|---|---|---|---|
| https://api.makeyourlifeeasier.web.id/v1 | 2 | 0 | 0% | openai/gpt-5 (1) |
| https://rhz4zlv.abc-tunnel.us/v1 | 1 | 0 | 0% | deepseek/deepseek-v4-flash (1) |
| https://ongoing-separated-transcription-realty.trycloudflare.com/v1 | 1 | 78 | 0% | — |
| https://arrives-knife-operation-carlos.trycloudflare.com/v1 | 1 | 75 | 0% | — |
Recent anomalies
https://api.makeyourlifeeasier.web.id/v1Aug 21, 2026, 9:59 PM
- Self-claimed model
- df-model
- V2 predicted family
- deepseek
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://api.makeyourlifeeasier.web.id/v1Aug 21, 2026, 9:05 PM
- Self-claimed model
- df-model
- V2 predicted family
- openai
- V3 best match
- openai/gpt-5 (sim 0.85)
- Verdict
- Full swap
https://rhz4zlv.abc-tunnel.us/v1Aug 18, 2026, 4:37 AM
- Self-claimed model
- qd/dfmodel
- V2 predicted family
- deepseek
- V3 best match
- deepseek/deepseek-v4-flash (sim 0.75)
- Verdict
- Full swap
https://ongoing-separated-transcription-realty.trycloudflare.com/v1Aug 18, 2026, 4:02 AM
- Self-claimed model
- qd/dfmodel
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://arrives-knife-operation-carlos.trycloudflare.com/v1Aug 18, 2026, 3:37 AM
- Self-claimed model
- qd/dfmodel
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
Methodology
BazaarLink Probe combines a Bayesian V2 family classifier across nine fingerprint categories with V3 sub-model references to verify endpoints claiming dfmodel. Low family consistency, a different V3 match, or divergence from selfClaim is flagged as anomalous. BazaarLink Probe home.