How does pricing work?

You pay only for what you use. Each model has separate input, output, and cache token rates; we publish the Aonis price, official API reference, and discount for every public model.

Do I need separate accounts with each AI provider?

No. One account gives you access to multiple supported models through a single API and a single balance.

Can I switch between models?

Yes. Choose the model that fits each workload and switch whenever you need without maintaining separate provider integrations.

Is the API ready for production use?

Yes. Aonis provides documented public model IDs, per-token pricing, and one OpenAI-compatible endpoint. Review the documented model and integration support before shipping to production.

Which models are available?

The public catalog lists current model IDs, token prices, official API comparisons, and supported integrations in one place.

How do I get started?

Create an account, add balance, generate an API key, choose a model, and send your first request. No separate provider setup required.

[ 01 / 09 ]

[ HERO ]

YOUR UNIFIED MULTI-MODEL AI API

One API. Multiple Models.Lower Costs.

RunClaude

Connect the models you need with one key, one balance, and public pricing. Save up to 85% compared with official API rates.

One API. Leading AI models.

Kimi

Z.AI

One API connects you to the models that matter. Switch between GPT, Claude, DeepSeek, Gemini and more without changing your stack.

[ 02 / 09 ]

[ FEATURES ]

ONE ENDPOINT. MULTIPLE MODELS. ZERO FRICTION.

One API. Every Model You
Need.

Use the models you need through one OpenAI-compatible endpoint. Compare public token rates with official API pricing before you build.

POST /v1/chat/completions

01 import OpenAI from "openai";
02 
03 const client = new OpenAI({
04   baseURL: "https://api.aonis.space/v1",
05   apiKey: process.env.AONIS_API_KEY,
06 });
07 
08 const completion = await client.chat.completions.create({
09   model: "claude-sonnet-5", // select a model below10   messages: [
11     { role: "system", content: "You are a helpful assistant." },
12     { role: "user", content: "Explain quantum computing in simple terms." },
13   ],
14   temperature: 0.7,
15 });
16 
17 console.log(completion.choices[0].message.content);

Switch models. Not your stack. Claude, GPT, DeepSeek, Grok, Gemini and more through the same integration.

[ 03 / 09 ]

[ BENEFITS ]

WHY DEVELOPERS CHOOSE US

Build once. Choose freely.

Aonis gives your product one stable foundation, while the model behind each request stays yours to choose.

One endpoint.

Make one integration and route every request through the same API, key and billing layer.

api.aonis.ai/v1

Every leading model.

Move between Claude, GPT, Gemini, DeepSeek and more without touching the foundation of your app.

CLAUDE / GPT / GEMINI / DEEPSEEK

No stack migration.

Stay OpenAI-compatible, preserve your workflow and choose the model that fits each moment.

COMPATIBLE BY DEFAULT

Add balance. Create a key. Pick a model. Send your first request.

[ 04 / 09 ]

[ HOW TO USE ]

HOW IT WORKS

From signup to your first request in three simple steps.

01/01
Add balance
Create your account and add credit to your balance. No subscription or fixed plan required.
02/02
Create your API key
Generate a key, choose the model you want, and connect using our OpenAI-compatible endpoint.
03/03
Start building
Send your first request and track usage, tokens, and spending from your dashboard.

No subscriptions. No fixed plans. Just usage-based pricing.

[ 05 / 09 ]

[ PRICING ]

PAY AS YOU GO

Model catalog / token pricing

Simple pricing. Pay only for what you use.

Compare Aonis public rates with official API prices, choose the right model, and keep one balance across your stack.

19 models shown / USD per 1M tokens

Model / providerInput / 1MOutput / 1MCache read / 1MCache write / 1M

Claude Fable 5claude-fable-5 / AnthropicOfficial API: $10.00 / $50.0085% off

$1.50

$7.50

$0.15

$1.875

Claude Opus 5claude-opus-5 / AnthropicOfficial API: $5.00 / $25.0085% off

$0.75

$3.75

$0.075

$0.9375

Claude Opus 4.8claude-opus-4-8 / AnthropicOfficial API: $5.00 / $25.0085% off

$0.75

$3.75

$0.075

$0.9375

Claude Opus 4.7claude-opus-4-7 / AnthropicOfficial API: $5.00 / $25.0085% off

$0.75

$3.75

$0.075

$0.9375

Claude Sonnet 5claude-sonnet-5 / AnthropicOfficial API: $2.00 / $10.0085% off

$0.30

$1.50

$0.03

$0.375

Claude Sonnet 4.6claude-sonnet-4-6 / AnthropicOfficial API: $3.00 / $15.0085% off

$0.45

$2.25

$0.045

$0.5625

GPT-5.6 Solgpt-5.6 / OpenAIOfficial API: $5.00 / $30.0085% offAliases: gpt-5.6-sol, gpt-5.6-fast, codex-5.6-sol

$0.75

$4.50

$0.075

$0.9375

GPT-5.6 Sol Ultragpt-5.6-sol-ultra / OpenAIOfficial API: $5.00 / $30.0085% offAliases: codex-5.6-sol-ultra

$0.75

$4.50

$0.075

$0.9375

GPT-5.6 Terragpt-5.6-terra / OpenAIOfficial API: $2.00 / $12.0085% offAliases: codex-5.6-terra

$0.30

$1.80

$0.03

$0.375

GPT-5.6 Lunagpt-5.6-luna / OpenAIOfficial API: $0.20 / $1.2075% offAliases: codex-5.6-luna

$0.05

$0.30

$0.005

$0.0625

Gemini 3.1 Progemini-3.1-pro / GoogleOfficial API: $2.00 / $12.0080% off

$0.40

$2.40

$0.04

$0.40

Gemini 2.5 Progemini-2.5-pro / GoogleOfficial API: $1.25 / $10.0080% off

$0.25

$2.00

$0.025

$0.25

DeepSeek V4 Prodeepseek-v4-pro / DeepSeekOfficial API: $0.435 / $0.8740% off

$0.261

$0.522

$0.005695

$0.261

DeepSeek V4 Flashdeepseek-v4-flash / DeepSeekOfficial API: $0.14 / $0.2825% off

$0.105

$0.21

$0.00427125

$0.105

Qwen 3.7 Maxqwen3.7-max / QwenOfficial API: $1.475 / $4.42580% off

$0.295

$0.885

$0.0295

$0.3688

Qwen 3.6 Plusqwen3.6-plus / QwenOfficial API: $0.325 / $1.9565% off

$0.11375

$0.6825

$0.011375

$0.142205

Kimi K3kimi-k3 / Moonshot AIOfficial API: $3.00 / $15.0080% off

$0.60

$3.00

$0.06

$0.60

Grok 4.5grok-4.5 / xAIOfficial API: $2.00 / $6.0080% off

$0.40

$1.20

$0.06

N/A

GLM-5.2glm-5.2 / Z.AIOfficial API: $1.40 / $4.4080% off

$0.28

$0.88

$0.052

$0.28

Transparent pricing across every model. Compare costs, choose what fits your workload, and switch models whenever you need.

[ 06 / 09 ]

[ TESTIMONIALS ]

REAL FEEDBACK FROM BUILDERS

Built for developers. Loved by teams.

Ethan Moore
@ethanmoore
I was tired of maintaining separate integrations for every provider. Now I can test different models through the same API and switch without touching half the codebase.
Lucas Bennett
@lucasbennett
What sold me was how little I had to change. I already had an OpenAI-style integration, changed the endpoint and model, and kept building.
Olivia Tran
@oliviatran
We don't need the most expensive model for every request. Being able to use different models for different workloads from one place makes a lot more sense.
Elena Roque
@elenaroque
Before this, testing another provider meant another account, another key, another billing setup, and another integration. Now it's much easier to compare options.
Maya Singh
@mayasingh
The API feels predictable, which is exactly what I want from infrastructure. I can try a new model without redesigning the application around another provider.
James Carter
@jamescarter
I started using it just to compare models. It ended up being the simplest way to keep multiple providers available without maintaining multiple implementations.
Ethan Moore
@ethanmoore
I was tired of maintaining separate integrations for every provider. Now I can test different models through the same API and switch without touching half the codebase.
Lucas Bennett
@lucasbennett
What sold me was how little I had to change. I already had an OpenAI-style integration, changed the endpoint and model, and kept building.
Olivia Tran
@oliviatran
We don't need the most expensive model for every request. Being able to use different models for different workloads from one place makes a lot more sense.
Elena Roque
@elenaroque
Before this, testing another provider meant another account, another key, another billing setup, and another integration. Now it's much easier to compare options.
Maya Singh
@mayasingh
The API feels predictable, which is exactly what I want from infrastructure. I can try a new model without redesigning the application around another provider.
James Carter
@jamescarter
I started using it just to compare models. It ended up being the simplest way to keep multiple providers available without maintaining multiple implementations.

Ethan Moore
@ethanmoore
I was tired of maintaining separate integrations for every provider. Now I can test different models through the same API and switch without touching half the codebase.
Lucas Bennett
@lucasbennett
What sold me was how little I had to change. I already had an OpenAI-style integration, changed the endpoint and model, and kept building.
Olivia Tran
@oliviatran
We don't need the most expensive model for every request. Being able to use different models for different workloads from one place makes a lot more sense.
Elena Roque
@elenaroque
Before this, testing another provider meant another account, another key, another billing setup, and another integration. Now it's much easier to compare options.
Maya Singh
@mayasingh
The API feels predictable, which is exactly what I want from infrastructure. I can try a new model without redesigning the application around another provider.
James Carter
@jamescarter
I started using it just to compare models. It ended up being the simplest way to keep multiple providers available without maintaining multiple implementations.
Ethan Moore
@ethanmoore
I was tired of maintaining separate integrations for every provider. Now I can test different models through the same API and switch without touching half the codebase.
Lucas Bennett
@lucasbennett
What sold me was how little I had to change. I already had an OpenAI-style integration, changed the endpoint and model, and kept building.
Olivia Tran
@oliviatran
We don't need the most expensive model for every request. Being able to use different models for different workloads from one place makes a lot more sense.
Elena Roque
@elenaroque
Before this, testing another provider meant another account, another key, another billing setup, and another integration. Now it's much easier to compare options.
Maya Singh
@mayasingh
The API feels predictable, which is exactly what I want from infrastructure. I can try a new model without redesigning the application around another provider.
James Carter
@jamescarter
I started using it just to compare models. It ended up being the simplest way to keep multiple providers available without maintaining multiple implementations.

Great infrastructure should stay out of the way. One API, consistent behavior, and the freedom to choose the right model for every workload.

[ 07 / 09 ]

[ FAQS ]

ANSWERS TO YOUR COMMON QUESTIONS

Got questions? We've got answers.

You pay only for what you use. Each model has separate input, output, and cache token rates; we publish the Aonis price, official API reference, and discount for every public model.

One API, clear pricing, and access to the models you need. Everything is designed to keep your infrastructure simple as you build and scale.

[ 08 / 09 ]

[ CTA ]

GET STARTED TODAY

Ready to build with the models you need?

Go from account to first request in minutes, with public model pricing and one OpenAI-compatible endpoint.

Start building View API docs

One API | Multiple models | Lower costs | Built for developers

One API. Multiple Models.Lower Costs.

One API. Every Model You Need.

Build once. Choose freely.

One endpoint.

Every leading model.

No stack migration.

From signup to your first request in three simple steps.

Add balance

Create your API key

Start building

Simple pricing. Pay only for what you use.

Built for developers. Loved by teams.

Got questions? We've got answers.

Ready to build with the models you need?

One API. Every Model You
Need.