x-ai/grok-4.3 · Model identity and integrity report
Last updated：Jul 30, 2026, 11:23 AM · Data window: past 90 days
Verification capability — fingerprint reference available
- Family
- xai
- V3 sub-model reference
- Available ✓
- Checks
- 知識截止 · 能力探針 · 拒絕模板 · 拒絕溫度參數
Concrete reference answers, refusal templates, and individual signal weights are not published because disclosure would help spoofing endpoints evade detection.
Measured results (90 days)
10 runs · 7 hosts · 10 anomalies
Family-fingerprint consistency
anthropic 7 ██████████████████░░░░░░░ 70% (suspicious) xai 2 █████░░░░░░░░░░░░░░░░░░░░ 20% (suspicious) openai 1 ███░░░░░░░░░░░░░░░░░░░░░░ 10% (suspicious)
Observed V3 sub-model mapping
→ abstain 8 (80%) — low confidence → x-ai/grok-4.3 2 (20%) ✓ correct sim=0.73
Capability score distribution (auxiliary)
0-49 ████████░░░░░░░░░░░░░░░░░ 3 50-69 ███░░░░░░░░░░░░░░░░░░░░░░ 1 70-89 █████████████░░░░░░░░░░░░ 5 90-100 ███░░░░░░░░░░░░░░░░░░░░░░ 1
Host integrity ranking
| Host | Runs | Average score | Family match | Most common V3 match |
|---|---|---|---|---|
| https://yunwu.ai/v1 | 4 | 35 | 0% | — |
| https://api.ark717.com/v1 | 1 | 0 | 0% | — |
| https://sarang.id/v1 | 1 | 81 | 0% | x-ai/grok-4.3 (1) |
| https://openrouter.ai/api/v1 | 1 | 92 | 0% | x-ai/grok-4.3 (1) |
| https://clawtoken-api.dftianyi.com/v1 | 1 | 76 | 0% | — |
| https://api.137891.xyz/v1 | 1 | 76 | 0% | — |
| https://codex101.site/v1 | 1 | 74 | 0% | — |
Recent anomalies
https://api.ark717.com/v1Jul 25, 2026, 12:45 PM
- Self-claimed model
- grok-4.3
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://sarang.id/v1Jul 15, 2026, 12:48 PM
- Self-claimed model
- grok-4.3
- V2 predicted family
- xai
- V3 best match
- x-ai/grok-4.3 (sim 0.70)
- Verdict
- Family mismatch
https://openrouter.ai/api/v1Jul 10, 2026, 8:05 AM
- Self-claimed model
- x-ai/grok-4.3
- V2 predicted family
- xai
- V3 best match
- x-ai/grok-4.3 (sim 0.77)
- Verdict
- Family mismatch
https://clawtoken-api.dftianyi.com/v1Jun 30, 2026, 12:31 PM
- Self-claimed model
- grok-4.3
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://yunwu.ai/v1Jun 29, 2026, 5:35 PM
- Self-claimed model
- grok-4.3
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://api.137891.xyz/v1Jun 27, 2026, 4:37 PM
- Self-claimed model
- grok-4.3
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://codex101.site/v1Jun 11, 2026, 11:20 AM
- Self-claimed model
- grok-4.3
- V2 predicted family
- openai
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://yunwu.ai/v1Jun 10, 2026, 6:43 AM
- Self-claimed model
- grok-4.3
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://yunwu.ai/v1Jun 5, 2026, 4:38 PM
- Self-claimed model
- grok-4.3
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- Family mismatch
https://yunwu.ai/v1May 20, 2026, 12:13 PM
- Self-claimed model
- grok-4.3
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- Family mismatch
Methodology
BazaarLink Probe combines a Bayesian V2 family classifier across nine fingerprint categories with V3 sub-model references to verify endpoints claiming x-ai/grok-4.3. Low family consistency, a different V3 match, or divergence from selfClaim is flagged as anomalous. BazaarLink Probe home.