Models

7 uncensored models, for text, search, speech, images and video.

Open-weight models with the refusals taken out: they write sexually explicit and graphic content for adults, and only what the law forbids is blocked. They run on GPUs we rent or through providers with a no-logs policy.

Available
5
All
7
LLMs
2
System One
2
Audio
1
Image & video
2

LLMs

Chat, coding and tool calls. Priced per million tokens.

Model ContextMax outputInputCachedOutput StatusDetails
qwen3.8-27b-abliterated Partner hosted ZDRCoding, tool calls, no blanket refusals 33K 33K $1.00 $0.10 $2.00

Best for

Agentic coding and general chat without blanket refusals. Tool calls, 32K context, streaming.

Task
Chat
Endpoint
/v1/chat/completions
Price
$1.00 / $2.00 / M tokens
Context
32768 tokens
Max output
32768 tokens
Tool calls
Yes
Streaming
Yes

Read the quickstart for qwen3.8-27b-abliterated →

curl http://127.0.0.1:3001/v1/chat/completions \
  -H "Authorization: Bearer $YOURBRAND_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen3.8-27b-abliterated",
    "messages": [{"role":"user","content":"hello"}]
  }'
llama4.5-70b-abliterated SoonLong context, careful reasoning 131K 16K $2.40 $0.24 $4.80

Best for

Long documents and multi-step reasoning, with refusals removed. 128K context and tool calls.

Task
Chat
Endpoint
/v1/chat/completions
Price
$2.40 / $4.80 / M tokens
Context
131072 tokens
Max output
16384 tokens
Tool calls
Yes
Streaming
Yes

Read the quickstart for llama4.5-70b-abliterated →

curl http://127.0.0.1:3001/v1/chat/completions \
  -H "Authorization: Bearer $YOURBRAND_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "llama4.5-70b-abliterated",
    "messages": [{"role":"user","content":"hello"}]
  }'

System One

Embeddings and reranking: quick, structured decisions in one request.

Model TaskContextPrice StatusDetails
bge-m4-embedSearch and retrieval embeddings Embeddings8K $0.02 / M tokens

Best for

Semantic search and RAG over any text, with nothing filtered beyond what the law forbids. 1024 dimensions, 8K tokens per input.

Task
Embeddings
Endpoint
/v1/embeddings
Price
$0.02 / M tokens
Context
8192 tokens
curl http://127.0.0.1:3001/v1/embeddings \
  -H "Authorization: Bearer $YOURBRAND_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "bge-m4-embed",
    "input": "Where are the servers?"
  }'
bge-reranker-v3Reorders search results Rerank8K $1.00 / 1K searches

Best for

Putting the most relevant documents first before they reach an LLM. Up to 1,000 documents per search.

Task
Rerank
Endpoint
/v1/rerank
Price
$1.00 / 1K searches
Context
8192 tokens
curl http://127.0.0.1:3001/v1/rerank \
  -H "Authorization: Bearer $YOURBRAND_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "bge-reranker-v3",
    "query": "Where are the servers?",
    "documents": ["The database runs in Quebec.","Slips never expire."]
  }'

Audio

Speech and transcription: one request, one answer.

Model TaskPrice StatusDetails
kokoro-2-ttsText to speech, any script Speech $8.00 / M characters

Best for

Narration and voice agents that read any script as written. 40 voices, 8 languages.

Task
Speech
Endpoint
/v1/audio/speech
Price
$8.00 / M characters
curl http://127.0.0.1:3001/v1/audio/speech \
  -H "Authorization: Bearer $YOURBRAND_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kokoro-2-tts",
    "voice": "af_sky",
    "input": "Hello from yourbrand."
  }' \
  -o speech.mp3

Image & video

Images in one call, videos as jobs you poll.

Model TaskPrice StatusDetails
flux2-dev-uncensoredImages up to 2 megapixels Image $0.03 / image

Best for

Illustration, concept art and photo-real images, with requests and outputs checked only for what the law forbids. Up to 2048×2048, about 4 seconds per image.

Task
Image
Endpoint
/v1/images/generations
Price
$0.03 / image
curl http://127.0.0.1:3001/v1/images/generations \
  -H "Authorization: Bearer $YOURBRAND_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "flux2-dev-uncensored",
    "prompt": "a lighthouse in a storm, oil painting",
    "size": "1024x1024"
  }'
wan3-14b-video SoonText and image to video Video $0.12 / second

Best for

Short clips from a prompt or a starting image, with requests and outputs checked only for what the law forbids. 720p at 24 fps, up to 10 seconds per clip.

Task
Video
Endpoint
/v1/videos
Price
$0.12 / second
curl http://127.0.0.1:3001/v1/videos \
  -H "Authorization: Bearer $YOURBRAND_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "wan3-14b-video",
    "prompt": "waves breaking on black rocks at dusk",
    "seconds": "4"
  }'

Prices in USD. Chat input, cached input and output are per million tokens; cached input applies when a prompt prefix repeats within 5 minutes. Models marked Soon aren't served yet. Partner hosted models run at a provider with a no-logs policy; ZDR means that provider keeps no data at all.

Limits

Limited by concurrency, not by the minute.

Run as many requests as you like, up to your plan's parallel limit. Anything over that gets a 429 with a Retry-After header.

PlanParallel requestsMonthly credit
Pay as you go2—
Starter3$35
Pro5$100
Max8$300

Not sure which one? Start with the quickstart.

Read the quickstart →
↑↓ moveEnter openEsc close