Bonus: Top up now and we'll double your first deposit — get x2 credits instantly.

White label

Buy inference wholesale.
Sell it at your price.

Partner rates up to 30% below list on 400+ models. One endpoint with your brand.

Trusted by 2100+ companies globally

The model

You resell our router as your own product.

Take the platform.

Your logo, your domain, your prices.

Give it to your customers.

They sign up, get keys, and use it as your product.

You pay the partner rate.

You charge your price. You keep the difference on every call.

Nothing to build. The providers are contracted, the models are integrated, the platform is running today.

What your customers get

An application, not just an endpoint.

Dashboards

Spend, tokens, latency and errors — per model, per key, per customer.

Guardrails

Rules on which models, providers and regions each key can reach.

EvalLab

Run the same test cases across models, compare output and cost side by side.

API key management

Keys with hard spend caps and daily or monthly resets.

Team management

Organizations, projects, and owner / admin / developer roles.

Admin panel

Your side of the platform: every account, its usage, its pricing, its limits.

30+ providers

Contracted and integrated. OpenAI, Anthropic, Google, AWS Bedrock, Azure, xAI, Mistral, Groq, Cerebras, DeepSeek, Alibaba, Together and more.

400+ models

Text, vision, image, video, audio and embeddings under one API.

All of it on your domain, under your brand.

Build vs buy

Three things you skip.

The build.

The full feature set already exists and is in production use by 2100+ companies. Nothing to design, ship or QA.

The maintenance.

We run the platform, staff the support, and add new models and providers as they launch. Your engineers stay on your own roadmap.

The paperwork.

SOC 2, ISO 27001, GDPR, zero data retention by default. You can sell into every region from day one instead of writing a compliance programme first.

The catalogue

400+ models across 30+ providers. One API.

List prices per 1M tokens. Your partner rate sits up to 30% below these — your customers see whatever you set above it.

ModelProviderInput / 1MOutput / 1MContext
Claude Opus 5Anthropic$5.00$25.001M
GPT-6 AstraOpenAI$10.00$50.001.05M
GPT-5.6 TerraOpenAI$2.00$12.001.05M
Gemini 3.8 FlashGoogle$0.75$3.751.05M
Grok 4.6xAI$2.00$6.00500K
Kimi K3Moonshot$3.00$15.001.05M
Qwen 3.8 MaxAlibaba$2.00$6.001M
DeepSeek V4.1 FlashDeepSeek$0.3$1.201M
GLM-5.3Z.ai$1.40$4.401.05M
Mistral Large 3Mistral$0.5$1.50262K

See all 400+ models →

Text generation Reasoning Vision and image input Image generation Video generation Speech-to-text Text-to-speech Embeddings and reranking Tool calling Streaming OCR and documents

Whatever your customers ask for next, it is already in the catalogue.

Setup

Three steps.

01Brand it.

Logo, colours, favicon, your domain. No code.

02Price it.

Set your own rate per model or per account. Change prices without a redeploy.

03Earn the spread.

You invoice your customers on your terms. We invoice you once a month for the partner rate. We never contact your customers.

Partners

Already running under someone else's brand.

2100+companies on the platform
400+models in the catalogue
30+contracted providers

Bring your volume. Get your rate.

Thirty minutes on your model mix and your numbers.