What is an AI API Gateway?

An AI API Gateway is a unified API endpoint that provides access to multiple AI models (like Claude, GPT, Gemini) through a single interface. It handles routing, billing, rate limiting, and fallbacks automatically.

Is Osiris compatible with the OpenAI SDK?

Yes. Osiris is fully OpenAI SDK compatible. Change your base URL to point to Osiris and use any of the 200+ supported models without changing your code.

How much does Osiris cost?

Osiris uses pay-per-use pricing in USD. You only pay for the tokens you use. There are no monthly fees or minimum charges. New users get free trial credits.

Which AI models does Osiris support?

Osiris supports 200+ AI models including Claude, GPT, Gemini, DeepSeek, Llama, and more from providers like OpenAI, Anthropic, Google, Meta, and others. All accessible through a single API endpoint.

Can I use Osiris with the Anthropic SDK?

Yes. Osiris supports both OpenAI and Anthropic SDK formats. You can call any model through either SDK without changing your existing code.

$ AI ROUTING INFRASTRUCTURE

One precise gateway for every AI workload.

Ship with Claude, GPT, Gemini, DeepSeek and more through OpenAI, Anthropic, or Responses APIs. One key, observable usage, resilient routing.

Start building →Read the docs

request.route● LIVE

$ curl /v1/chat/completions --model osiris-claude-sonnet-4.6 → auth ........ verified → policy ...... allowed → route ....... provider-01 ← stream ...... connected

OpenAI-compatibleREADY

Anthropic MessagesREADY

Responses APIREADY

AVAILABLE MODEL FAMILIES

ClaudeAnthropic

GPTOpenAI

GeminiGoogle

GLMZ.AI

DeepSeekDeepSeek

QwenAlibaba

KimiMoonshot

GrokxAI

MiniMaxMiniMax

3 API formats

Usage-based billing

Automatic fallback

Streaming ready

$ ACCESS MODES

Use the gateway your way.

Choose the resource model that matches your workload. The API remains the same.

MODE_01

Credit

Pay only for actual token usage. Credits never expire on new purchases.

MODE_02

Request

Predictable daily, weekly, or monthly request allowances.

MODE_03

Token

Weighted token pools for high-throughput development workloads.

MODE_04

Time

Unlimited access for a defined active period.

$ QUICKSTART

From key to first response in minutes.

Create a key

Choose a plan and generate one secure API key.

Change the base URL

Keep your existing SDK and point it to Osiris.

Route any model

Select a model ID; routing and fallback happen automatically.

$ PRICING

Models & Plans

Per 1M token pricing and credit packages. Loaded live from active configuration.

Anthropic

10 models

ModelInput / 1MOutput / 1MCache / 1M

Claude Fable 5VISION

InputRp 8.5K

OutputRp 42.7K

CacheRp 2.1K

Claude Fable 5.1VISION

InputRp 8.5K

OutputRp 42.7K

CacheRp 2.1K

Claude Opus 4.7VISION

InputRp 4.3K

OutputRp 21.3K

CacheRp 1.1K

Claude Opus 4.8VISION

InputRp 4.3K

OutputRp 21.3K

CacheRp 1.1K

Claude Opus 5VISION

InputRp 4.3K

OutputRp 21.3K

CacheRp 1.1K

Claude Sonnet 4.5VISION

InputRp 2.3K

OutputRp 11.4K

Cache—

Claude Sonnet 5VISION

InputRp 3.0K

OutputRp 14.9K

CacheRp 746

Claude-Haiku-4.5VISION

InputRp 853

OutputRp 4.3K

CacheRp 213

Claude-Opus-4.6VISION

InputRp 4.3K

OutputRp 21.3K

CacheRp 1.1K

Claude-Sonnet-4.6VISION

InputRp 2.6K

OutputRp 12.8K

CacheRp 640

OpenAI

5 models

ModelInput / 1MOutput / 1MCache / 1M

GPT 5.5VISION

InputRp 2.8K

OutputRp 14.2K

CacheRp 711

GPT 6 AstraVISION

InputRp 8.5K

OutputRp 42.7K

CacheRp 2.1K

GPT-5.6 LunaVISION

InputRp 1.6K

OutputRp 8.0K

CacheRp 400

GPT-5.6 SolVISION

InputRp 2.8K

OutputRp 14.2K

CacheRp 711

GPT-5.6 TerraVISION

InputRp 2.1K

OutputRp 10.7K

CacheRp 533

Google

1 model

ModelInput / 1MOutput / 1MCache / 1M

Gemini-3.1-Pro

InputRp 6.5K

OutputRp 43.1K

Cache—

Zhipu

11 models

ModelInput / 1MOutput / 1MCache / 1M

GLM 4.6

InputRp 1.1K

OutputRp 4.2K

CacheRp 212

GLM 4.6VVISION

InputRp 531

OutputRp 1.6K

CacheRp 106

GLM 4.7VISION

InputRp 1.6K

OutputRp 7.4K

CacheRp 318

GLM 5.0 Turbo

InputRp 1.4K

OutputRp 6.8K

CacheRp 341

GLM 5.1VISION

InputRp 1.6K

OutputRp 8.0K

CacheRp 398

GLM 5.3VISION

InputRp 1.6K

OutputRp 8.0K

CacheRp 398

GLM 5.3 FlashVISION

InputRp 0

OutputRp 0

CacheRp 0

GLM 5.3 FlashXVISION

InputRp 0

OutputRp 0

CacheRp 0

GLM 5V TurboVISION

InputRp 1.4K

OutputRp 6.8K

CacheRp 341

GLM-5.0VISION

InputRp 1.1K

OutputRp 5.7K

CacheRp 284

glm-5.2VISION

InputRp 1.6K

OutputRp 8.0K

CacheRp 398

DeepSeek

2 models

ModelInput / 1MOutput / 1MCache / 1M

DeepSeek V4 ProVISION

InputRp 1.4K

OutputRp 7.1K

CacheRp 355

DeepSeek V4.1 FlashVISION

InputRp 0

OutputRp 0

CacheRp 0

Alibaba

6 models

ModelInput / 1MOutput / 1MCache / 1M

Qwen 3.7 MaxVISION

InputRp 1.3K

OutputRp 6.3K

CacheRp 315

Qwen 3.7 PlusVISION

InputRp 637

OutputRp 3.2K

CacheRp 159

Qwen 3.8 MaxVISION

InputRp 7.2K

OutputRp 21.6K

Cache—

Qwen3.8 27BVISION

InputRp 0

OutputRp 0

CacheRp 0

Qwen3.8 FlashVISION

InputRp 0

OutputRp 0

CacheRp 0

Qwen3.8-Max-PreviewVISION

InputRp 1.7K

OutputRp 8.5K

CacheRp 427

Moonshot

5 models

ModelInput / 1MOutput / 1MCache / 1M

Kimi K2.5VISION

InputRp 648

OutputRp 3.2K

CacheRp 162

Kimi K2.6VISION

InputRp 1.1K

OutputRp 5.4K

CacheRp 270

Kimi K2.7VISION

InputRp 1.1K

OutputRp 5.4K

CacheRp 270

Kimi K2.8 PreviewVISION

InputRp 0

OutputRp 0

CacheRp 3

Kimi K3VISION

InputRp 2.6K

OutputRp 12.8K

CacheRp 640

xAI

2 models

ModelInput / 1MOutput / 1MCache / 1M

Grok 4.5VISION

InputRp 1.7K

OutputRp 8.5K

CacheRp 427

Grok 4.6VISION

InputRp 1.7K

OutputRp 8.5K

CacheRp 142

MiniMax

3 models

ModelInput / 1MOutput / 1MCache / 1M

MiniMax M2.5VISION

InputRp 1.1K

OutputRp 4.3K

CacheRp 108

MiniMax M2.7

InputRp 0

OutputRp 0

CacheRp 0

MiniMax M3VISION

InputRp 597

OutputRp 3.0K

CacheRp 149

Tencent

5 models

ModelInput / 1MOutput / 1MCache / 1M

Hunyuan 2.0 Instruct

InputRp 503

OutputRp 2.0K

Cache—

Hunyuan 3 Preview

InputRp 1.3K

OutputRp 6.4K

CacheRp 320

Hunyuan 3 Preview Agent

InputRp 647

OutputRp 2.2K

Cache—

Hunyuan Chat

InputRp 503

OutputRp 2.0K

Cache—

hy3

InputRp 0

OutputRp 0

CacheRp 0

Muse AI

2 models

ModelInput / 1MOutput / 1MCache / 1M

Muse Spark 1.1

InputRp 1.4K

OutputRp 7.1K

CacheRp 355

Muse Spark 1.2

InputRp 1.4K

OutputRp 7.1K

CacheRp 355

Other

5 models

ModelInput / 1MOutput / 1MCache / 1M

Hunyuan 4 Preview

InputRp 0

OutputRp 0

CacheRp 0

Xiaomi MiMo v2.6 FlashVISION

InputRp 0

OutputRp 0

CacheRp 0

Xiaomi MiMo v2.6 ProVISION

InputRp 0

OutputRp 0

CacheRp 0

mimo-v2.5VISION

InputRp 0

OutputRp 0

CacheRp 0

mimo-v2.5-pro

InputRp 0

OutputRp 0

CacheRp 0

Credit Packages

Plan	Harga	Credits
Starter	Rp 89.852	Rp 89.852
Pro+20%	Rp 179.704	Rp 215.645
Max+25%	Rp 359.409	Rp 449.261
Enterprise+25%	Rp 718.817	Rp 898.522

Your next model is one request away.

Build against one stable API while the model landscape keeps moving.

Start building →