At a glance
Record- Operator
- AbliteratedLLM
- Group
- Dedicated lab
- KYC
- No KYC
- API shape
- OpenAI + Anthropic
- Founded
- Not published
- Website
- abliteratedllm.dev
AbliteratedLLM is listed as a dedicated lab provider with 18 models in the cybersec directory.
Crypto-only and prepaid. One abliterated model, GLM-5.3 Flash; 15 stock models and 2 official DeepSeek API models keep vendor refusals. Official DeepSeek prompts are stored in China. 18+.
No prompts or outputs stored on its own systems. Abliterated and stock models run under zero-data-retention. Official DeepSeek API models are processed and stored in China and may be used for training.
Source: abliteratedllm.dev/pricing ↗ · Updated Nov 7, 2026
OpenAI + Anthropic
No KYC
CyberBench 67.9
List rates. Cached input shown where published; "-" means the provider does not publish a cache rate. Workload comparison uses 1M input plus 250K output. Per 1M tokens. Cached input from $0.0054. Priority tier adds 25-50% on some stock models.
Crypto only via NOWPayments: Monero, Bitcoin, USDT, Ethereum, Litecoin and other coins. $5 free trial credit. First deposit of $10 or more gets a 20% bonus. No refunds. Paid limits: 600 RPM per key, 1,200 RPM per account, 64 concurrent. Trial: 10 RPM per key, 1 concurrent, 4,096 output tokens.
No KYC or email. Accounts are random 16-digit numbers. Crypto checkout via NOWPayments, which may run its own AML checks. 18+ only.
No ID or email at signup; anonymous account number + crypto payment
Vendor source ↗ · Basis: Vendor statement
Median and p95 response times for short and long prompts, measured from repeatable probes rather than vendor claims.
Coming soonStreaming output rate and request throughput under steady load, with concurrency and context length recorded.
Coming soonBlended cost for a fixed workload, including cache reads and output tokens, so speed and price can be compared together.
Coming soonIndependent performance probes are being added. Vendor-reported numbers are not mixed into this section.