Use the Anthropic SDK with abliteration.ai

How to use abliteration.ai with the Anthropic Messages API — base URL, auth token, and what’s supported.

To use the Anthropic SDK with abliteration.ai, set the base URL to https://api.abliteration.ai and pass your ak_... key as api_key= (sent as x-api-key) — no other code changes. abliteration.ai implements the Anthropic Messages API, including streaming, tool use, web search, and web fetch.

Configuration

Base URL	https://api.abliteration.ai
Auth	Authorization: Bearer ak_... or x-api-key: ak_... — both work
Endpoint	/v1/messages

Python

from anthropic import Anthropic

client = Anthropic(
    base_url="https://api.abliteration.ai",
    api_key=os.environ["ABLIT_KEY"],
)

resp = client.messages.create(
    model="abliterated-model",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}],
)

curl

curl https://api.abliteration.ai/v1/messages \
  -H "x-api-key: $ABLIT_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "abliterated-model",
    "max_tokens": 1024,
    "messages": [{ "role": "user", "content": "Hello" }]
  }'

Policies

Published policies scoped to your project apply automatically. Pass policy_ids to add published policies selected for this request. See Policies in API requests.

Prompt caching

Prompt-prefix caching is automatic. Anthropic-style cache_control objects are accepted for SDK compatibility, but they do not guarantee a cache hit or a fixed retention period on abliteration.ai. Use a stable x-abliteration-session-id header to improve routing affinity, then check usage.cache_read_input_tokens to confirm a hit. See prompt caching.

Reasoning

All models reason before answering. On Messages, control it with:

• output_config: { "effort": "high" } — set the reasoning depth (low … max).
• thinking: { "type": "enabled", "budget_tokens": N } — extended thinking with a token budget, mapped to an effort level. If this and output_config.effort are set, output_config.effort wins.
• thinking: false — disable reasoning on the base or previous large model. Large V2 instead runs at low and hides the trace. This Large V2 compatibility behavior also overrides an explicit output_config.effort.

To keep reasoning but hide the trace: include_reasoning: false removes the thinking blocks from the response, and thinking.display: "omitted" returns a thinking block with its text emptied.

The reasoning trace is returned as a thinking content block. See thinking & reasoning effort for per-model behavior.

abliterated-model-large and abliterated-model-large-v2 are text-only. Requests with image content are rejected with a 400 — send images to abliterated-model.