One API.
Every AI.
Connect once, get access to a smart router that judges your request and sends it to the best AI — automatically, at the lowest possible cost.
OpenAI Compatible
Works with Cursor
Works with Hermes
Claude Code ready
Streaming support
Spend limits
Start with the API
Why Diva Router?
One key. Smart routing. Lowest cost. No lock-in.
🧠
Smart Judge
Every request is classified — simple chat goes cheap, hard reasoning goes strong. You never overpay.
🔀
Auto Failover
If one provider is down, requests instantly route to the next available model. Zero downtime.
🔑
Your Own Keys
Each customer gets a virtual key with a spend budget. Provider keys stay private to you.
💸
Cheapest Lane
Free tier for simple tasks, ultra-fast lane for chat, premium long-context models. Auto-selected.
📡
OpenAI Protocol
Drop-in replacement for any app using the OpenAI SDK. Change one URL, keep your code.
📊
Usage Dashboard
See every request, token count, and cost per customer. Set hard budgets to stop runaway bills.
Model Lanes
Pick a lane, or use diva-auto and let the smart router choose the cheapest healthy provider.
diva-auto RECOMMENDED
Adaptive routing by price, latency, live load, failures, and available quota
auto · cheapest
diva-chat
General chat & Q&A — fast, low-cost lane
low cost
diva-long
Long documents, 1M context
1M context
diva-reason
Math, logic, coding — strongest reasoning
strong
diva-vision
Image understanding — attach an image and ask
multimodal
diva-voice TTS
Text-to-speech — natural AI voice for your apps
audio out
diva-transcribe STT
Speech-to-text — voice messages, meetings, audio files
audio in
diva-free
Free lane — unlimited simple requests ($0)
$0 cost
Choose Your Plan
Subscription, token packs, or pay-as-you-go — pick what fits. No hidden costs.
📅 Subscription — daily token quota
| Plan | Price | Daily token quota | Access |
|---|
| Free | $0 | 100K tokens/day | diva-free lane |
| Starter | $9/mo | 1M tokens/day | + diva-chat |
| Pro | $29/mo | 5M tokens/day | + diva-long (capped) |
| Business | $99/mo | 25M tokens/day | + flagship lanes |
| Enterprise | $299/mo | 100M tokens/day | all lanes, custom |
🎟️ Token Packs — buy tokens, use anytime
| Pack | Price | Tokens |
|---|
| Micro | $2 | 1M tokens |
| Small | $15 | 10M tokens |
| Medium | $60 | 50M tokens |
| Large | $250 | 250M tokens |
⚡ Pay-As-You-Go — only pay for what you use
# No subscription. No expiry. Top up and go.
Base URL = https://ai.litellam.tech/v1
Model = diva-auto
# You are billed per token used, at pack rates.
Quickstart
Works with any OpenAI-compatible client. Change one URL, keep everything else.
1
Get your API key — contact us for a key + your spend budget.
2
Point your app at the router:
BASE_URL = https://ai.litellam.tech/v1
API_KEY = sk-customer-xxxx
MODEL = diva-auto
cURL
curl https://ai.litellam.tech/v1/chat/completions \
-H "Authorization: Bearer sk-customer-xxxx" \
-H "Content-Type: application/json" \
-d '{"model":"diva-auto","messages":[{"role":"user","content":"Hello!"}]}'
Python (OpenAI SDK)
from openai import OpenAI
client = OpenAI(base_url="https://ai.litellam.tech/v1", api_key="sk-customer-xxxx")
r = client.chat.completions.create(
model="diva-auto",
messages=[{"role": "user", "content": "Explain quantum computing"}]
)
print(r.choices[0].message.content)
Cursor IDE
Settings → Models → OpenAI API Key:
Base URL: https://ai.litellam.tech/v1
API Key: sk-customer-xxxx
Model: diva-auto
Usage & Limits
Fair use, transparent metering.
⏱️ Rate limits
Default: 5 req/s per key, 100k tokens/min. Raise on request.
💰 Budgets
Hard spend cap per key. Requests stop automatically at the limit.
📈 Analytics
Per-key dashboard: requests, tokens, cost, latency. Exportable.
🔒 Privacy
Your prompts are not used for training. Provider keys never exposed.