One API for every AI model.
Point the OpenAI SDK at LazySusan and use GPT-6, Claude, Gemini, Grok, DeepSeek, Perplexity, FLUX, Nano Banana and Veo with one key. Switching models is a one-string change.
- 18 chat · 13 image · 6 video models
- no subscription
- credits from $25
from openai import OpenAI client = OpenAI( base_url="https://www.lazysusan.ai/api/v1", api_key="lsk_live_…",) reply = client.chat.completions.create( model="anthropic/claude-sonnet-5-5", messages=[{"role": "user", "content": "Draft a reply."}],)change one line, switch the model
anthropic/claude-sonnet-5-5
Claude Sonnet 5.5
openai/gpt-6.1-sol
GPT-6.1 Sol
google/gemini-3.8-flash
Gemini 3.8 Flash
what you get
Wire it up once. Use any model after.
Stop keeping a separate account, key, SDK and invoice for every provider. One integration covers chat, images and video.
one key
Every model behind one key.
One account, one balance and one bill for every provider. No separate accounts with OpenAI, Anthropic, Google and the rest.
openai-compatible
Keep the SDK you already use.
The OpenAI SDKs for Python and Node work as they are. Change the base URL and the key, and keep the rest of your code.
18 chat models
Streaming, tools and vision.
Chat Completions with streaming, function calling, JSON output and image input, across every chat model.
13 image models
Images from one endpoint.
Nano Banana, GPT Image, FLUX and Grok Imagine through /images/generations, returned as a hosted URL or base64.
6 video models
Video as a job you poll.
Start a Veo 3.1 or Runway clip, check its status, collect the file. Failed videos are refunded in full.
usage per request
Every call shows its cost.
Responses include the tokens used and the credits charged. The console lists every request.
spend controls
Limits you set.
A monthly spend cap per account, keys you can revoke in one click, and rate limits per key.
auto top-up
A balance that refills itself.
Buy credits once, then let the balance top up from your saved card whenever it drops below your threshold.
how it works
From sign-in to first response in a few minutes.
step 1
Create a key
Sign in to the console and create a key. It's shown once, so store it somewhere safe.
step 2
Add credits
Buy a pack from $25, or turn on auto top-up after your first purchase.
step 3
Call any model
Send requests to the base URL below with the model id you want. Change the id to change the model.
// curl
curl https://www.lazysusan.ai/api/v1/chat/completions \
-H "Authorization: Bearer $LAZYSUSAN_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "google/gemini-3.8-flash", "messages": [{"role": "user", "content": "Hello"}]}'credits
Prepaid credits. No subscription.
Buy a pack and spend it on any model. 1 credit = $0.001.
$25
25,000 credits
$100
100,000 credits
$500
500,000 credits
$2,000
2,000,000 credits
Charged for what you use
Each request reserves the most it could cost, then you're charged only what it used. The rest is released right away.
Failures are free
If a provider errors or a video doesn't render, the reservation is returned in full.
Auto top-up, if you want it
Pick a threshold and an amount. When the balance drops below it, your saved card is charged and credits are added.
Credits don't expire. Custom amounts from $25 to $5,000 in the console.
teams and enterprise
Rolling AI out across a company?
One vendor instead of six, one invoice instead of six, and one place to see what every team spends.
or email support@lazysusan.ai
volume pricing
Lower rates on committed monthly spend.
invoicing
Pay by invoice instead of card on larger commitments.
higher limits
More than the default 120 requests a minute per key, when you need it.
a named contact
Someone who answers when your integration has a question.
keys stored hashed
We keep only a hash of each key. The full key is shown once, at creation.
no prompt storage
We don't store your prompts or outputs, and we don't train models on them.
rate card
Every model, priced in credits.
1 credit = $0.001. Chat is priced per million tokens, images per image and video per clip. The same list is available from GET /api/v1/models.
chat · credits per 1M tokens
| Model id | Name | Input | Cached input | Output |
|---|---|---|---|---|
| openai/gpt-6.1-sol | GPT-6.1 Sol | 2,300 | 115 | 11,500 |
| openai/gpt-6-astra | GPT-6 Astra | 11,500 | 1,150 | 57,500 |
| openai/gpt-6-luna | GPT-6 Luna | 115 | 11.5 | 575 |
| anthropic/claude-fable-5-1 | Claude Fable 5.1 | 11,500 | 11,500 | 57,500 |
| anthropic/claude-opus-5-5 | Claude Opus 5.5 | 4,600 | 4,600 | 23,000 |
| anthropic/claude-sonnet-5-5 | Claude Sonnet 5.5 | 2,300 | 2,300 | 11,500 |
| anthropic/claude-haiku-5-5 | Claude Haiku 5.5 | 115 | 115 | 575 |
| google/gemini-3.1-pro | Gemini 3.1 Pro | 2,300 | 2,300 | 13,800 |
| google/gemini-3.8-flash | Gemini 3.8 Flash | 862.5 | 862.5 | 4,312.5 |
| google/gemini-3.5-flash-lite | Gemini 3.5 Flash-Lite | 345 | 345 | 2,875 |
| xai/grok-4.7 | Grok 4.7 | 2,300 | 575 | 6,900 |
| xai/grok-4.3 | Grok 4.3 | 1,437.5 | 230 | 2,875 |
| xai/grok-build-0.1 | Grok Build | 1,150 | 230 | 2,300 |
| deepseek/deepseek-flash | DeepSeek V4.1 Flash | 345 | 6.9 | 1,380 |
| deepseek/deepseek-v4-pro | DeepSeek V4 Pro | 1,518 | 50.6 | 4,554 |
| perplexity/sonar | Sonarplus web search, billed as used | 2,300 | 2,300 | 2,300 |
| perplexity/sonar-pro | Sonar Proplus web search, billed as used | 6,900 | 6,900 | 34,500 |
| perplexity/sonar-reasoning-pro | Sonar Reasoning Proplus web search, billed as used | 4,600 | 4,600 | 18,400 |
image · credits per image
| Model id | Name | Per image |
|---|---|---|
| google/nano-banana-2.1 | Nano Banana 2.1 | 38.64 |
| google/nano-banana-pro | Nano Banana Pro | 154.1 |
| openai/gpt-image-2.5 | GPT Image 2.5billed by tokens used, never more than shown | billed by usage · typically 5.64 (low) to 48.42 (high) |
| openai/gpt-image-2.5-fast | GPT Image 2.5 Fastbilled by tokens used, never more than shown | billed by usage · typically 5.64 (low) to 48.42 (high) |
| bfl/flux-2-pro | FLUX 2 Pro | 57.5 |
| bfl/flux-3 | FLUX 3 | 55.2 |
| bfl/flux-kontext-pro | FLUX Kontext Pro | 46 |
| bfl/flux-kontext-max | FLUX Kontext Max | 92 |
| bfl/flux-pro-1.1 | FLUX 1.1 Pro | 46 |
| bfl/flux-pro-1.1-ultra | FLUX 1.1 Ultra | 69 |
| xai/grok-imagine-image | Grok Imagine | 23 |
| xai/grok-imagine-image-quality | Grok Imagine Quality | 57.5 |
| xai/grok-imagine-image-2.0 | Grok Imagine 2.0 | 69 |
video · credits per clip
| Model id | Name | Per clip |
|---|---|---|
| google/veo-3.1 | Veo 3.1with audio | 720p 4s: 1,840 · 720p 6s: 2,760 · 720p 8s: 3,680 · 1080p 8s: 3,680 |
| google/veo-3.1-fast | Veo 3.1 Fastwith audio | 720p 4s: 460 · 720p 6s: 690 · 720p 8s: 920 · 1080p 8s: 1,104 |
| google/veo-3.1-lite | Veo 3.1 Litewith audio | 720p 4s: 230 · 720p 6s: 345 · 720p 8s: 460 · 1080p 8s: 736 |
| runway/gen-4.5 | Runway Gen-4.5 | 720p 5s: 690 · 720p 10s: 1,380 |
| runway/gen-4-turbo | Runway Gen-4 Turbo | 720p 5s: 287.5 · 720p 10s: 575 |
| minimax/hailuo-2.3 | Hailuo 2.3 | 768p 6s: 322 · 768p 10s: 644 · 1080p 6s: 563.5 |
Yes for /chat/completions, /images/generations and /models: same request and response shapes, so the OpenAI SDKs work after you change the base URL and key. Video uses its own small async endpoint (/videos), because the OpenAI SDK has no equivalent.
Anything that talks to the OpenAI API with a custom base URL: the official OpenAI SDKs for Python and Node, LangChain's ChatOpenAI, the Vercel AI SDK's OpenAI-compatible provider, or plain HTTP from any language.
From a prepaid credit balance (1 credit = $0.001). Chat is priced per token, images per image and video per clip; the rate card on this page lists every model. Before a request runs we reserve the most it could cost, then charge only what it actually used and release the rest straight away.
Requests that the balance can't cover are refused with a 402 error before anything runs, so you are never billed past zero. Turn on auto top-up if you'd rather the balance refill itself.
No. There's no subscription and no monthly minimum. Credits stay on your account until you use them.
We don't train models on your data, and we don't store your prompts or outputs. We keep request records (model, token counts, credits charged) for billing. Generated images and videos are hosted for you at an unlisted URL that is only returned to you.
Same login, separate balance. App plans and app tokens cover the workspace; the API has its own prepaid credits, so the two never mix.
120 requests a minute per key and 600 per account by default. Need more? Email support@lazysusan.ai.
get started
One key. Every model on the menu.
Create a key, add $25 of credits and send your first request. Your code stays the same when you switch models.