Uncensored · Preparing launch
Uncensored DeepSeek V4.1 Flash API
DeepSeek V4.1 Flash with modified uncensored weights, one of five models on Lyber's launch lineup. Once it is qualified, call it by its model ID through the same OpenAI-compatible API, API key and prepaid credit as every other model.
Pick a named checkpoint or send to lyber/auto. Every answer tells you which model produced it.
Model facts
- Model ID
lyber/deepseek41-flash-abliterated- Weights
- Uncensored
- Launch target
- $0.30 / $1.20USD per 1 million tokens · fresh input / outputCached input: $0.0075 per 1M tokens
- Hardware plan
- 1017 checkpoint target · mixed FP8/FP4
- Serving plan
- dealign.ai’s uncensored 1017 checkpoint is in qualification. Hosted checkpoint identity and serving terms are checked separately.
- Status
- Preparing launch
Inference is unavailable until each model and its hardware are qualified. Launch targets are provisional; the API's published price version governs billing.
Call it with an OpenAI SDK.
Set the base URL to Lyber and pass this model ID. The response's model field names the model that answered.
import os
from openai import OpenAI
client = OpenAI(
base_url="https://lyber-paid-acceptance-gateway-production.up.railway.app/v1",
api_key=os.environ["LYBER_API_KEY"],
)
reply = client.chat.completions.create(
model="lyber/deepseek41-flash-abliterated",
messages=[
{"role": "user", "content": "Answer what I asked."},
],
)
print(reply.model)
print(reply.choices[0].message.content)We work directly with model researchers and verify modified weights, commercial hosting rights, features and completed answers. Each served model’s identity stays explicit. We never substitute base GLM or Kimi.