OpenAI- and Anthropic-compatible API

Every major AI model. One API key.

Anthropic, OpenAI, Google, xAI, DeepSeek, Mistral, and Moonshot AI models, plus speech, transcription, image and video, through one endpoint your SDK already speaks. Pay as you go at the providers' list prices.

Get your API key Read the docs

No subscription
Automatic fallback
Invoice for every top-up

curl https://api.sparrowcodes.com/v1/chat/completions \
  -H "Authorization: Bearer $SPARROW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "anthropic/claude-sonnet-5.5",
    "messages": [{"role": "user", "content": "Explain rate limiting in one paragraph."}]
  }'

Example response200 OK

x-sparrow-provideranthropic

x-sparrow-modelanthropic/claude-sonnet-5.5

x-sparrow-cost-usd0.001676

usage18 in · 164 out

Models from

Anthropic
OpenAI
Google
xAI
DeepSeek
Mistral
Moonshot AI

23 chat models, 4 more model types

How it works

What happens to every request

Your code sends one request. SparrowCodes checks it, applies your privacy and routing rules, and keeps trying providers until one answers.

Your app

POST /v1/chat/completions
{
  "model": "anthropic/claude-sonnet-5.5",
  "models": ["openai/gpt-6.1-sol"],
  "provider": { "sort": "price" },
  "messages": [ … ]
}

Your normal SDK call, with an optional backup model and routing preference.

SparrowCodes

Key, limits and balance
Rejected before any provider is called if a limit would be passed.
Privacy rules
Personal data masked if you turned masking on. Nothing stored by default.
Cache
An identical recent request is answered instantly, at no token cost.
Routing
Providers ordered by your preferences and their live health.

Providers

Anthropic429 rate limited
anthropic/claude-sonnet-5.5Not charged
OpenAI200 answered
openai/gpt-6.1-solCharged at list price

Each attempt is logged. You see which provider answered and exactly what it cost.

Why SparrowCodes

Built for production traffic

The same models and the same API format. The difference is everything around the request.

Automatic fallback: List backup models in a request. If a provider errors or rate-limits, the next one answers, and you pay only for the answer.
Routing you control: Pin a provider order, allow or exclude providers, sort by price or speed, and cap the price you're willing to pay per request.
Response cache: Identical requests can be answered from cache in milliseconds at no token cost. Every cached response is marked in its headers.
Zero retention: Prompts and answers are not stored unless you turn logging on. Optional masking strips emails, phone numbers and card numbers before they leave.
Bring your own keys: Already have a provider account? Add its key and keep your own rates and limits, with SparrowCodes routing, logs and fallback on top.
Teams and roles: Owner, admin, developer and billing roles, two-factor sign-in, and an audit log of every change to keys, members and settings.
Spend limits: Monthly limits per key and alerts before your balance runs low, so a runaway script or a leaked key always has a ceiling.
Every request, itemised: Tokens, latency, provider, each fallback attempt and the exact cost, searchable and exportable to CSV.
Webhooks and Slack: Get low-balance, limit and key events in your own systems or a Slack channel as they happen.

Drop-in

Works with the SDKs you already use

Point your client at SparrowCodes by changing the base URL and the key. Streaming, tools, images and structured output work the way your SDK expects.

OpenAI SDK for Python
OpenAI SDK for Node.js
Anthropic SDK for Python
Vercel AI SDK

Setup guides

Endpoint	What it does
/v1/chat/completions	Chat, tools, vision, streaming. OpenAI format.
/v1/messages	Claude's native format. Works with the Anthropic SDK.
/v1/responses	OpenAI Responses format, stateless.
/v1/embeddings	Text embeddings for search and RAG.
/v1/images/generations	Image generation.
/v1/videos	Video generation, polled until ready.
/v1/audio/speech	Text to speech.
/v1/audio/transcriptions	Speech to text.

Models

Provider list prices, per million tokens

No markup on tokens: you pay what the provider charges. The top model from each provider is below; the full catalog has 23 chat models, 5 image models, 3 video models, 5 transcription models, and 2 embedding models.

All models

Model	Provider	Context	Input / 1M	Output / 1M	Supports
Claude Fable 5.1 anthropic/claude-fable-5.1	Anthropic	1M	$10	$50	Tools, Images
GPT-6 Astra openai/gpt-6-astra	OpenAI	1.05M	$10	$50	Tools
Gemini 3.1 Pro (preview) google/gemini-3.1-pro	Google	1.05M	$2	$12	Tools, Images
Grok 4.7 x-ai/grok-4.7	xAI	500K	$2	$6	Tools, Images
DeepSeek V4 Pro deepseek/deepseek-v4-pro	DeepSeek	1.05M	$1.32	$3.96	Tools
Mistral Large 4 mistral/mistral-large-4	Mistral	1.05M	$0.70	$2.10	Tools, Images
Kimi K3 moonshotai/kimi-k3	Moonshot AI	262K	$3	$15	Tools

Pricing

Prepaid credits. No plans.

You add credits and each request spends exactly what it uses, at the provider's list price. A 5.5% platform fee applies when you buy credits. That's the whole pricing model.

Credits added to your balance: $100.00
Platform fee (5.5%): $5.50
You pay: $105.50

Pricing calculator

Make your first request in minutes

Get your API key Browse models