API Documentation

The two things every tool asks for

Base URL: https://badassmodel.com/v1

API Key: sk-bam-... (Dashboard → API Keys → Create key)

curl (the 30-second test)

curl https://badassmodel.com/v1/chat/completions \
  -H "Authorization: Bearer sk-bam-..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-r1",
    "stream": true,
    "messages": [{"role": "user", "content": "hello"}]
  }'

List models: curl https://badassmodel.com/v1/models -H "Authorization: Bearer sk-bam-..."

Pick your model

Every model you can call is in GET /v1/models. Take the id of the one you want and put it in the "model" field of your request — that’s the whole selection mechanism, same as OpenAI.

curl https://badassmodel.com/v1/models -H "Authorization: Bearer sk-bam-..."

→ {"data": [
    {"id": "euryale-l3.3-70b-uncensored", "kind": "chat",  ...},
    {"id": "qwen3-vl-235b-instruct",      "kind": "chat",  ...},
    {"id": "flux-1-dev",                  "kind": "image", ...},
    {"id": "kokoro-82m",                  "kind": "tts",   ...},
    ...]}

The kind tells you which endpoint serves it: chat → /chat/completions, image → /images/generations,tts → /audio/speech. The live list with per-model prices is on the pricing page.

Vision: chat models with vl or vision in the name (e.g. qwen3-vl-235b-instruct) accept images via OpenAI content parts —{"type": "image_url", "image_url": {"url": "https://…"}} alongside your text.

Python (official openai SDK)

from openai import OpenAI

client = OpenAI(base_url="https://badassmodel.com/v1", api_key="sk-bam-...")

stream = client.chat.completions.create(
    model="deepseek-r1",
    messages=[{"role": "user", "content": "hello"}],
    stream=True,
)
for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="")

OpenCode

export BADASS_API_KEY=sk-bam-...
# add to ~/.config/opencode/opencode.json:
{
  "provider": {
    "badassmodeles": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "BadAssModeles",
      "options": { "baseURL": "https://badassmodel.com/v1", "apiKey": "{env:BADASS_API_KEY}" },
      "models": { "deepseek-r1": { "name": "DeepSeek R1" } }
    }
  }
}

Cursor / VS Code (Cline, Roo, Continue) / Open WebUI / SillyTavern

Choose OpenAI Compatible as the provider and fill in:

Base URL: https://badassmodel.com/v1

API Key: sk-bam-...

Model: any id from GET https://badassmodel.com/v1/models

Image generation

OpenAI-shaped POST /v1/images/generations. Billed flat per image returned (see pricing); n is capped at 4.

curl https://badassmodel.com/v1/images/generations \
  -H "Authorization: Bearer sk-bam-..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "flux-1-schnell",
    "prompt": "a cat riding a skateboard",
    "n": 1,
    "size": "1024x1024"
  }'

Returns {"created": ..., "data": [{"b64_json": "..."}]}.

Text to speech

OpenAI-shaped POST /v1/audio/speech. Billed per input character.

curl https://badassmodel.com/v1/audio/speech \
  -H "Authorization: Bearer sk-bam-..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kokoro-82m",
    "input": "The quick brown fox jumps over the lazy dog"
  }' \
  --output speech.wav

Returns raw audio bytes (audio/wav or audio/mpeg).

Billing & limits

Prepaid credits only — $1 = 1 credit. Deposit via crypto (NOWPayments, no fee) or Whop (card, Apple Pay, more — 5% processing fee, so a $50 deposit adds 47.50 credits); $5 minimum. Every request is billed from your balance — per token for chat, per image for generations, per character for speech — no balance, no requests (HTTP 402). 60 requests/min per key. Deposit any amount from $5 up; a single deposit of $50+, $100+, or $200+ permanently unlocks lower rates (see pricing). All deposits are final — no refunds.