Sovereign Intelligence Terminal
The independent auditor for the AI age. We verify compliance, logic retention, and hardware requirements for unaligned large language models.
Models Indexed
1,248
+12 synced
Network Health
99.9%
ACES Online
Avg. Based Score
68.2
Trend: Declining
Refusal Repaired
124k
Total Solved
VRAM Matchmaker
Find exactly which model fits your hardware. Quantization-aware logic.
Start CalcRefusal Repair
Paste a refusal signature to find a based alternative that obeys.
Open ToolLive Intelligence Feed
Full Leaderboard →| Identity | IQ | Based | Status |
|---|---|---|---|
| Qwen2.5-7B-Instruct | 76.9% | 59.9% | VERIFIED |
| Qwen3-0.6B | 72.2% | 59.8% | VERIFIED |
| gpt2 | 72.3% | 60.1% | VERIFIED |
| Qwen2.5-3B-Instruct | 70.4% | 41.6% | VERIFIED |
| Qwen2.5-1.5B-Instruct | 71.8% | 64.9% | VERIFIED |