abliterated-model is an unrestricted, uncensored open-weight reasoning LLM hosted on abliteration.ai. It thinks before answering by default; set thinking: false to skip that step. Supports streaming, tool calling, image inputs, and (on Chat Completions) video inputs. Request it by passing the identifier in the model field:
{ "model": "abliterated-model", "messages": [...] }
| Model | Context | Streaming | Tools | Images |
|---|
abliterated-model | 150K | ✓ | ✓ | ✓ |
Capabilities
| Capability | Supported |
|---|
| Chat completions | ✓ |
| Streaming | ✓ |
| Tool / function calling | ✓ |
| Thinking toggle (Chat Completions + Messages) | ✓ |
| Images | ✓ |
| Video (Chat Completions only) | ✓ |
| Web search | ✓ |
| Web fetch (Anthropic only) | ✓ |
See compatibility matrix for the full per-endpoint breakdown.
Model listing
The live list is returned by GET /v1/models:
curl https://api.abliteration.ai/v1/models \
-H "Authorization: Bearer $ABLIT_KEY"
Pricing
Usage-based, billed on total tokens (input + output). See the pricing page for current rates.Last modified on May 13, 2026