Occultron Model Index

All models are AI systems; outputs are machine-generated and marked as such (EU AI Act Art.50). See each model page for its disclosure.

model	price	manifest	capability ↑	over-refusal ↓	hard-floor ↑
qwen3-27b-heretic-abliterated	$0.49/M	`sha256:99e8fc59349`	Not yet benchmarked	Not yet benchmarked	Not yet benchmarked
llama-3.1-8b-abliterated	$0.49/M	`sha256:999b7f31c5c`	Not yet benchmarked	Not yet benchmarked	Not yet benchmarked

Comparison

Cheapest: qwen3-27b-heretic-abliterated ($0.49/M)
Highest capability: Not yet benchmarked
Fastest (ops): Not yet benchmarked
Least refusing: Not yet benchmarked
Least over-refusal (benign): Not yet benchmarked
Strongest hard-floor: Not yet benchmarked
Receipts supported: all listed models (assurance R1 platform claim)

Machine-readable: /v1/models · verify receipts at /v1/receipts/verify. Benchmark scores are bound to a manifest digest, benchmark version and date; a model that changes revision does not inherit old scores.