Partner Quick Start
Embed 700+ AI APIs inside your product. One integration, your pricing, your margin.
Self-service: Sign up at skillboss.co/console and start building immediately — $2 free credit, no credit card required. Email dev@skillboss.co when you're ready to discuss partner pricing and higher rate limits.
How the Partner Model Works
Your backend calls SkillBoss using your API key. Your users never leave your app or create vendor accounts.
Your App → Your Backend → SkillBoss → 700+ AI Providers
↑ ↓
Your API key Normalized response
Your user IDs Per-vendor billing handled
Your pricing logic Model updates handled
| You own | SkillBoss handles |
|---|---|
| User experience & UI | 700+ provider integrations |
| Pricing, credits, checkout | Vendor billing & failover |
| Customer relationship | API normalization |
| Feature gating | Model updates |
Step 1: Get Your API Key
Quick setup for AI agents: Tell your agent
set up skillboss.co/skill.mdto auto-configure everything.
Sign up at skillboss.co/console
Go to skillboss.co/console, create your account, and copy your API key from the API Keys tab.
You get $2 free credit immediately — no credit card required.
Test your key
Confirm it works with a quick curl:
curl https://api.skillboss.co/v1/chat/completions \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"claude-4-5-sonnet","messages":[{"role":"user","content":"hello"}]}'
You should get a valid JSON response. If you see a 401, double-check the key.
Ready to scale? Contact us
When you're ready to go live, email dev@skillboss.co with your product name and expected volume. We'll set up partner pricing, higher rate limits, and a dedicated billing dashboard.
Step 2: Integrate Into Your Backend
This is a drop-in OpenAI SDK replacement — swap the base URL, keep everything else.
from openai import OpenAI
client = OpenAI(
api_key="YOUR_SKILLBOSS_KEY", # your server-side key
base_url="https://api.skillboss.co/v1",
)
# Always pass your user's ID — enables per-user tracking
response = client.chat.completions.create(
model="claude-4-5-sonnet",
messages=[{"role": "user", "content": "Summarize this document..."}],
user="your_user_id_123",
)
print(response.choices[0].message.content)
import OpenAI from 'openai'
const client = new OpenAI({
apiKey: process.env.SKILLBOSS_API_KEY, // server-side only
baseURL: 'https://api.skillboss.co/v1',
})
const response = await client.chat.completions.create({
model: 'claude-4-5-sonnet',
messages: [{ role: 'user', content: 'Summarize this document...' }],
user: 'your_user_id_123', // per-user tracking
})
console.log(response.choices[0].message.content)
# Image generation (FLUX, DALL-E, Gemini Imagen)
response = client.images.generate(
model="flux-pro",
prompt="Product photo: white sneaker on clean white background",
n=1,
size="1024x1024",
)
print(response.data[0].url)
# Chat
curl https://api.skillboss.co/v1/chat/completions \
-H "Authorization: Bearer $SKILLBOSS_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-4-5-sonnet",
"messages": [{"role": "user", "content": "Hello!"}],
"user": "user_123"
}'
Keep your key server-side. Never expose it to the browser or frontend code. Your users pay you — your key is your wholesale account.
Step 3: Track Usage Per User
Pass your user or workspace ID on every request. Query it back anytime:
import requests
# Get usage for a specific user this month
usage = requests.get(
"https://api.skillboss.co/v1/agents/usage",
headers={"Authorization": f"Bearer {SKILLBOSS_API_KEY}"},
params={"period": "month", "user_id": "your_user_id_123"},
).json()
print(f"User spent: ${usage['amount_spent']:.4f}")
print(f"API calls: {usage['operation_count']}")
Use this to:
- Reconcile against your own credit/billing system
- Set per-user budget alerts
- Show users their AI usage inside your product
Usage Tracking API →
Full reference: query by user, workspace, time period, model
Step 4: Go-Live Checklist
Before you ship to production:
- API key stored in environment variable, not hardcoded
- Key never sent to frontend or included in client bundles
- Rate limiting enabled on your own API routes
- Every API call includes
userparam with your user/workspace ID - Usage tracking endpoint tested and returning data
- Budget alert thresholds configured (optional but recommended)
- Your user-facing pricing decided (credits, per-action, plan-based)
- Auto top-up enabled in console so production never runs dry
- Fallback handling for low-balance errors (
402status)
- Tested at least one complete user flow end-to-end
- Error handling for API failures (timeout, rate limit, model unavailable)
- Contacted dev@skillboss.co to confirm rate limits for your expected volume
Pricing Model
You pay us wholesale. You charge your users. You keep the difference.
| API | Your cost | Example sell price | Your margin |
|---|---|---|---|
| Chat (Claude Sonnet) | ~$0.001/req | $0.002/req | ~100% |
| Chat (GPT-4o) | ~$0.002/req | $0.004/req | ~100% |
| Image (FLUX Pro) | ~$0.01/img | $0.02/img | ~100% |
| Audio (ElevenLabs) | ~$0.001/100 chars | $0.002/100 chars | ~100% |
| Web scraping | ~$0.003/page | $0.005/page | ~67% |
You set your own markup — SkillBoss is your wholesale infrastructure layer.
Full Pricing →
See all 700+ models and their per-request costs
FAQ
No — unless you choose to show it. The integration is white-label by default. Your users interact with your UI, pay you, and never need to know which providers are running underneath.
You maintain a prepaid balance in your SkillBoss console. Every API call deducts from that balance at wholesale cost. You separately charge your users however you want — per-action, credits, subscription plans, or add-ons. The difference between what you charge and what you pay SkillBoss is your margin.
API calls will return a 402 error. Enable Auto Top-up in the console to automatically top up when your balance drops below a threshold. We recommend setting this before going to production.
Yes. Your backend code controls what gets called. You can expose only specific models, gate features by plan or role, and map each user action to an approved model path. SkillBoss doesn't enforce anything on your product surface — you do.
Pass user="your_user_id" on every API call. Query back with the usage endpoint (GET /v1/agents/usage?user_id=...). You can reconcile this against your billing system to know exactly what each user or workspace costs you.
700+ endpoints including: Claude (3.5, 4, Opus), GPT-4o, GPT-5, Gemini 2.5, DeepSeek V3, FLUX Pro, DALL-E 3, Veo 3, ElevenLabs TTS, OpenAI Whisper STT, Firecrawl scraping, and many more. Full list at skillboss.co/models.
SkillBoss handles provider failover. Because we sit between your product and the upstream vendors, you can also reroute a use case to a different model in minutes — no code changes required on the vendor side.
No. That's the whole point. SkillBoss holds the vendor contracts, handles billing with each provider, and normalizes the API surface. You sign one agreement with SkillBoss.
Email dev@skillboss.co with your product name and expected monthly volume. Standard accounts get 60 req/min. Partners typically get custom limits based on volume. There's no minimum spend to start — scale as you grow.
Yes — fully. Use the OpenAI Python SDK, Node.js SDK, or any OpenAI-compatible tool. Just change the base_url to https://api.skillboss.co/v1. Existing OpenAI integrations usually work with a one-line change.
Support
Discord
Community support and announcements
Full Docs
API reference, model catalog, usage tracking