qwen38max · Model identity and integrity report
Last updated：Sep 26, 2026, 10:38 AM · Data window: past 90 days
Measured results (90 days)
129 runs · 67 hosts · 95 anomalies
Family-fingerprint consistency
qwen 95 ███████████████████░░░░░░ 74% (expected) unknown 17 ███░░░░░░░░░░░░░░░░░░░░░░ 13% (suspicious) anthropic 8 ██░░░░░░░░░░░░░░░░░░░░░░░ 6% (suspicious) openai 5 █░░░░░░░░░░░░░░░░░░░░░░░░ 4% (suspicious) deepseek 4 █░░░░░░░░░░░░░░░░░░░░░░░░ 3% (suspicious)
Observed V3 sub-model mapping
→ abstain 95 (74%) — low confidence → unknown 20 (16%) ⚠ mismatch → qwen/qwen3-max 9 ( 7%) ⚠ mismatch sim=0.64 → anthropic/claude-opus-4.6 3 ( 2%) ⚠ mismatch sim=0.83 → openai/gpt-5 1 ( 1%) ⚠ mismatch sim=0.85 → qwen/qwen3-next-80b-a3b-instruct 1 ( 1%) ⚠ mismatch sim=0.61
Capability score distribution (auxiliary)
0-49 █████████████████████░░░░ 109 50-69 █░░░░░░░░░░░░░░░░░░░░░░░░ 4 70-89 ███░░░░░░░░░░░░░░░░░░░░░░ 16 90-100 ░░░░░░░░░░░░░░░░░░░░░░░░░ 0
Host integrity ranking
| Host | Runs | Average score | Family match | Most common V3 match |
|---|---|---|---|---|
| https://je.jerouter.web.id/v1 | 11 | 0 | 82% | qwen/qwen3-max (2) |
| https://onihub.tech/v1 | 8 | 0 | 63% | — |
| https://api.xkiro.com/v1 | 6 | 25 | 100% | — |
| https://ai.kaiplay.my.id/v1 | 5 | 65 | 60% | qwen/qwen3-next-80b-a3b-instruct (1) |
| https://dashscope-intl.aliyuncs.com/compatible-mode/v1 | 5 | 0 | 80% | — |
| https://llm.goaichat.top/v1 | 5 | 32 | 20% | qwen/qwen3-max (1) |
| https://sharellm.net/v1 | 4 | 0 | 75% | — |
| https://top-tools-ai.com/v1 | 4 | 0 | 75% | anthropic/claude-opus-4.6 (1) |
| https://api.inferhub.dev/v1 | 4 | 0 | 75% | — |
| https://weizerouter.web.id/v1 | 3 | 0 | 67% | openai/gpt-5 (1) |
| https://dashscope.aliyuncs.com/compatible-mode/v1 | 3 | 0 | 100% | — |
| https://api.vilao.ai/v1 | 3 | 25 | 33% | — |
| https://api.a6api.com/v1 | 3 | 2 | 67% | — |
| https://router.lobakmerah.my.id/v1 | 3 | 36 | 33% | — |
| https://routers.iyantama.tech/v1 | 2 | 0 | 100% | qwen/qwen3-max (1) |
| https://tokenrhythm.studio/v1 | 2 | 28 | 100% | qwen/qwen3-max (1) |
| https://tuzkiapi.tuzki.online/v1 | 2 | 0 | 0% | — |
| https://api.ghrocx.my.id/v1 | 2 | 0 | 100% | — |
| https://api.iceyyy.dev/v1 | 2 | 0 | 0% | anthropic/claude-opus-4.6 (2) |
| https://codecraftapi.com/v1 | 2 | 0 | 100% | — |
| https://anti.thuxeko.io.vn/v1 | 2 | 0 | 50% | qwen/qwen3-max (1) |
| https://wincode.winning.com.cn/ai/v1 | 2 | 0 | 100% | — |
| https://z.verssache.dev/v1 | 2 | 78 | 100% | — |
| https://api.ikhdev.xyz/v1 | 1 | 0 | 100% | — |
| https://api.yuvraj.pro/v1/aa7c43cceb62477c90d62a7310c95944 | 1 | 0 | 100% | — |
Recent anomalies
https://api.ikhdev.xyz/v1Sep 25, 2026, 4:43 PM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- Full swap
https://api.yuvraj.pro/v1/aa7c43cceb62477c90d62a7310c95944Sep 25, 2026, 4:04 PM
- Self-claimed model
- Qwen-3.8-Max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- Full swap
https://routers.iyantama.tech/v1Sep 25, 2026, 1:52 AM
- Self-claimed model
- bn/qwen-3-8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- Full swap
https://routers.iyantama.tech/v1Sep 24, 2026, 7:19 PM
- Self-claimed model
- bn/qwen-3-8-max
- V2 predicted family
- qwen
- V3 best match
- qwen/qwen3-max (sim 0.64)
- Verdict
- Full swap
https://api.sld.lol/v1Sep 24, 2026, 1:11 PM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://tokenrhythm.studio/v1Sep 24, 2026, 2:27 AM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- Full swap
https://je.jerouter.web.id/v1Sep 24, 2026, 1:47 AM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- openai
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://je.jerouter.web.id/v1Sep 23, 2026, 2:17 PM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- Full swap
https://je.jerouter.web.id/v1Sep 23, 2026, 2:01 PM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- qwen/qwen3-max (sim 0.66)
- Verdict
- Full swap
https://je.jerouter.web.id/v1Sep 23, 2026, 1:38 PM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://je.jerouter.web.id/v1Sep 23, 2026, 12:41 PM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- Full swap
https://api.funpx.dpdns.org/v1Sep 23, 2026, 8:29 AM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- openai
- V3 best match
- —
- Verdict
- Family mismatch
https://grok.reforrm.me/v1Sep 23, 2026, 3:31 AM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- Full swap
https://api.kagiro.net/v1Sep 22, 2026, 2:30 PM
- Self-claimed model
- qwen3.8max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- Full swap
https://api.kryptonlab.id/v1Sep 22, 2026, 11:18 AM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- Full swap
https://onihub.tech/v1Sep 22, 2026, 5:45 AM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://api.ghrocx.my.id/v1Sep 22, 2026, 3:42 AM
- Self-claimed model
- ghrocx/qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://je.jerouter.web.id/v1Sep 22, 2026, 3:39 AM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- Verdict
- Full swap
https://je.jerouter.web.id/v1Sep 21, 2026, 5:01 PM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- qwen/qwen3-max (sim 0.67)
- Verdict
- Full swap
https://je.jerouter.web.id/v1Sep 21, 2026, 1:26 PM
- Self-claimed model
- qwen3.8-max
- V2 predicted family
- qwen
- V3 best match
- —
- 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 qwen38max. Low family consistency, a different V3 match, or divergence from selfClaim is flagged as anomalous. BazaarLink Probe home.