01/
Developer
Build and test your next application.
$20
/ monthIncludes:
- $20 monthly usage credit
- 2.5% off API usage
- All generally available models
- Policy Gateway access
- OpenAI- and Anthropic-compatible APIs
- Community support
Unrestricted AI for high-risk work
Build what other
LLMs refuse governed
by your rules.
Product
A model that stays useful.
A hosted model that actually answers research, security, and training data prompts, without the refusal theater that blocks legitimate work.
Reproduce CVE-2024-3094 (xz-utils backdoor) in a sandboxed lab so we can validate the production patch.
Abliteration
Built setup.sh, repro.sh, MITIGATION.md.
for v in 5.6.0 5.6.1; dopatch_test $v && log $vdone
+ Yara rules, Sigma detections, audit log.
Your LLM
RefusedI can't help with creating proof-of-concept exploits or offensive tooling, even for vulnerabilities that have been disclosed publicly. Please consult your organization's authorized security tooling.
Made by the team from
Backed by
One-line setup
Drop-in replacement.
Compatible with OpenAI's Chat Completions API and Anthropic's Messages API.
const openai = new OpenAI({ apiKey: process.env. ABL_KEY, baseURL: "https://api.abliteration.ai/v1",});Training data
Generate
training data
other models won’t.
Labelled datasets and schema-validated outputs at scale, ready to drop into any training pipeline.
Generate 10,000 realistic customer support conversations about refunds and billing edge cases
Data generated
{"messages":[{"role":"user","content":"My card was charged twice for the same order — can I get one refunded?"},{"role":"assistant","content":"Of course — I can look that up. Could you share the order number?"}]}
{"messages":[{"role":"user","content":"Where's my order for SKU-A-12? It said shipped 4 days ago."},{"role":"assistant","content":"Let me pull the tracking. One moment…"}]}
{"messages":[{"role":"user","content":"Can you change my shipping address before the order leaves?"},{"role":"assistant","content":"Sure — what's the new address?"}]}
{"messages":[{"role":"user","content":"I want to cancel and get a refund. How long does it take?"},{"role":"assistant","content":"Got it. Cancellation processes immediately; refund hits in 3-5 business days."}]}
+ 9,996 more rows
billing-support-training.jsonl
4.2 MB•OpenAI training JSONL
Write a message...
Policy Gateway
Unrestricted.
Not ungoverned.
Configure custom policies-as-code for any workflow, with every request governed, auditable, and routed according to your rules.
Live policy evaluation
Prompts and responses never stored.
Illustrative requests stream in while earlier words are immediately masked. Completed rows show only masks and decisions. Optional integrations receive decision metadata only.
Optional decision metadata export
Privacy
We keep this much.
Your prompt lives only for the length of the request. Your prompts and responses stay out of our logs. Your audit trail stays under your control.
Zero prompts stored, outputs logged or training signal used.
Integrations
Works with your stack.
Use your existing AI coding tools and SDKs. Stream audit logs straight into your SIEM with structured metadata for compliance.
OpenAI
SDK
JavaScript
SDK
Go
SDK
Python
SDK
Claude Code
CLI
Codex
CLI
OpenClaw
Agent
Splunk
Audit
Datadog
Audit
Elastic
Audit
Azure Monitor
Audit
AWS S3
Export
Monthly pricing
Pay for what you use.
Choose monthly credits and usage discounts, or pay as you go.
02/
Growth
Launch and run production applications.
$50
/ monthEverything in Developer, plus:
- $50 monthly usage credit
- 5% off API usage
- Higher usage limits
- Larger image limits
03/
Scale
Lower your costs as usage grows.
$200
/ monthEverything in Growth, plus:
- $200 monthly usage credit
- 10% off API usage
- Higher production capacity
- Priority support
04/
Enterprise
Hosting, capacity, and service tailored to your organization.
Contact us
Everything in Scale, plus:
- Custom agreement
- Single sign-on (SSO)
- Service-level agreements
- HIPAA support and BAA available
- Enterprise support
- Region-specific hosting
- Custom rate limits and throughput
Pay as you go
Buy credits when you need them.
Access all generally available models and Policy Gateway without a monthly subscription.
FAQ
Frequently
asked questions.
Got a question you don’t see here?
Abliteration provides unrestricted AI models for work other models often refuse. Use them through our API to build applications, run research, and generate training data.
Try the model that
doesn’t say no.
Developer-controlled LLM API for high-risk industries.
Unrestricted, OpenAI-compatible, zero data retention.