99.99% uptime · 40+ models live

One API for every AI model your app needs.

BasedAPIs unifies LLMs, image generation, and embeddings behind a single, OpenAI-compatible endpoint. Swap models with one line — no SDK changes, no vendor lock-in.

Get your API key → View docs

quickstart.py

# pip install basedapis
from basedapis import BasedClient

client = BasedClient(api_key="ba_sk_...")

# Same call shape for every model
response = client.chat.completions.create(
    model="based-pro-1",
    messages=[
        {"role": "user", "content": "Explain zk-proofs in one line"}
    ]
)

print(response.choices[0].message.content)
# → "A zk-proof proves you know a secret without revealing the secret itself."

Why BasedAPIs

Built for builders who ship fast

Stop wiring separate SDKs, managing rate limits across five providers, and rewriting prompts every time you switch models. BasedAPIs handles all of it.

⚡

One endpoint, every model

OpenAI-compatible API. Switch from GPT-4o to Claude to open-weight models by changing one string. Zero code refactor.

🔐

Provable & verifiable

Every response ships with a cryptographic proof of inference. Verify model outputs on-chain or off — built on BasedAI's consensus network.

🛡️

Automatic failover

If a provider goes down, traffic reroutes in milliseconds. Your users never see a 500. We track uptime so you don't have to.

📊

Transparent usage & cost

Real-time dashboard with per-token, per-request cost breakdown. Set spending caps and alerts. No surprise bills.

🌐

Edge inference, global

Requests routed to the nearest inference node across 12 regions. Sub-200ms first-token latency for streaming completions.

📦

SDKs & native integrations

Python, TypeScript, Go, and Rust SDKs. LangChain, LlamaIndex, and Vercel AI SDK support out of the box. CLI for quick testing.

Model Library

40+ models, one API key

From frontier LLMs to fast, cheap open-weights. We add new models weekly — your integration never changes.

Based Pro 1

Based Mini

GPT-4o

GPT-4o mini

Claude 3.5 Sonnet

Claude 3 Haiku

Llama 3.1 405B

Llama 3.1 70B

Mistral Large 2

Mixtral 8x22B

Qwen 2.5 72B

DeepSeek V2

Grok-2

Gemini 1.5 Pro

Stable Diffusion 3

FLUX.1

DALL·E 3

Whisper Large

text-embedding-3

…and more →

Quickstart

Drop-in compatible. 3 lines to ship.

If you've used the OpenAI SDK, you already know BasedAPIs. Change one base URL — or install our SDK — and every model in the library is yours.

01 Install the SDK or point your existing OpenAI client at us.
02 Set base_url="https://api.basedapis.com/v1"
03 Swap the model string. That's it.

Python

cURL

TypeScript

from openai import OpenAI

client = OpenAI(
    base_url="https://api.basedapis.com/v1",
    api_key="ba_sk_..."
)

response = client.chat.completions.create(
    model="based-pro-1",
    messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)

# No SDK needed — plain cURL
curl https://api.basedapis.com/v1/chat/completions \
  -H "Authorization: Bearer ba_sk_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "based-pro-1",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.basedapis.com/v1",
  apiKey: "ba_sk_..."
});

const res = await client.chat.completions.create({
  model: "based-pro-1",
  messages: [{ role: "user", content: "Hello!" }]
});
console.log(res.choices[0].message.content);

Pricing

Pay for what you use

No seat fees, no platform fees. You pay per token at transparent rates — up to 80% cheaper than calling frontier models directly.

Developer

For prototypes & side projects

$0 /mo

10,000 tokens / day free
All open-weight models
Community support
Basic usage dashboard

Start free

Scale

For production apps

$29 /mo + usage

All 40+ models including frontier
Reduced per-token rates
Streaming & batch endpoints
Spending caps & alerts
Priority support

Get started

Enterprise

For teams at scale

Custom

Volume discounts on usage
Dedicated inference nodes
SSO, audit logs, SOC 2
99.99% uptime SLA
Solutions engineer

Contact sales