Apertis AI
OpenAI-compatible API gateway for 470 AI models
Last verified:
What is Apertis AI?
Apertis AI is an AI control plane designed specifically for coding agents and AI applications. It provides teams with a single OpenAI-compatible endpoint that aggregates access to 30+ AI model providers and 540+ models, including OpenAI, Anthropic, Google (Gemini + Vertex), AWS Bedrock, Mistral, DeepSeek, Alibaba (Qwen), Moonshot, Zhipu (GLM), xAI, Together, and Groq. The platform covers chat, embeddings, images, speech, and audio transcription capabilities.
Key features include virtual keys for issuing controlled API access per team or coding agent, budgets and model policy enforcement to keep usage within spending and model-access boundaries, intelligent routing and automatic failover across providers, and prompt caching to reduce repeated-context costs on coding workflows. Apertis offers drop-in integration where developers keep the OpenAI SDK and only change the base URL to https://api.apertis.ai/v1. The platform is an official provider for Vercel AI SDK, LiteLLM, LlamaIndex, and Kilo Code.
Apertis AI is built for development teams, engineering organizations, and anyone using coding agents like Claude Code, Cursor, Cline, Roo Code, or OpenCode. It is operated by Stima AI LLC, a US entity with enterprise-ready billing infrastructure including invoices, line-item receipts, and role-based team access. The platform provides per-request observability at apertis.ai/log to see which model, agent, and key was used in near real-time.
Apertis AI pricing
Pricing model: Freemium
Apertis offers two payment models that can be mixed. Subscription tiers include: Free at $0/mo (rate limited, perfect for testing and learning with limited API calls per day), Lite at $12/mo (monthly quota, higher rate limits than free, great for small projects), Pro at $25/mo (monthly quota, increased quota and priority support, ideal for growing applications), Plus at $60/mo (monthly quota, high quota with dedicated support, for production applications and teams), and Max at $180/mo. Pay-as-you-go lets you top up credits and pay per million tokens at per-model rates with no minimum spend. When a subscription quota is exhausted, requests fall back to PAYG credits with a hard cap you control. Enterprise plans offer custom pricing, SLA, dedicated support, and advanced features with custom quotas.
Apertis AI pros
- Access to 540+ AI models from 30+ providers through one API
- OpenAI-compatible API works as drop-in replacement with just base URL change
- Virtual keys enable controlled access per team, agent, or application
- Budget enforcement keeps agent usage within spending boundaries
- Model policy enforcement restricts which models agents can access
- Automatic routing and failover across multiple providers for reliability
- Free prompt cache reads reduce repeated-context costs on coding workflows
- Works with all major coding agents: Claude Code, Cursor, Cline, Roo Code, OpenCode
- Supports chat, embeddings, images, speech, and audio transcription endpoints
- Direct paid provider relationships ensure reliability won't hinge on free tiers
- Per-request observability dashboard shows model, agent, and key usage in real-time
- Official integrations with Vercel AI SDK, LiteLLM, LlamaIndex, and Kilo Code
- Flexible payment options: subscription tiers plus pay-as-you-go with no minimum
- Subscription quotas fall back to PAYG credits with controllable hard cap
- Free tier available for testing and learning with rate-limited API calls
- Enterprise plans offer custom pricing, SLA, and dedicated support
- No vendor lock-in since you can switch models with a single field change
Apertis AI cons
- Subscription pricing may be higher than direct provider rates for single-model users
- Pay-as-you-go costs can accumulate quickly without proper budget monitoring
- Free tier is rate-limited with limited API calls per day, not suitable for production
- Some model pricing varies significantly making cost prediction complex
- Requires changing base URL in existing code which may break some integrations
- Learning curve for managing virtual keys, budgets, and model policies
- Dependence on Apertis as intermediary adds one more point of potential failure
- Not all legacy models are supported compared to direct provider access
- Enterprise features require custom pricing negotiation rather than self-service
- Prompt cache benefits only apply to repeated-context workflows, not all use cases
Frequently asked questions about Apertis AI
What is Apertis?
Apertis is an AI control plane for coding agents and AI apps. It gives teams one OpenAI-compatible endpoint across 30+ AI model providers, plus virtual keys for issuing controlled access per team or agent, budgets and model policy to keep usage inside spending and model-access boundaries, routing and failover across providers, prompt cache to cut repeated-context cost on coding workflows, and drop-in integration — keep the OpenAI SDK and change only the base URL.
Which AI models and providers does Apertis support?
Apertis supports 30+ providers and hundreds of models across major regions: OpenAI, Anthropic, Google (Gemini + Vertex), AWS Bedrock, Mistral, DeepSeek, Alibaba (Qwen), Moonshot, Zhipu (GLM), xAI, Together, and Groq. Coverage includes chat, embeddings, images, speech, and audio transcription. The current model list with 540+ models is browsable at apertis.ai/models with pricing, context windows, and features.
Is Apertis OpenAI-compatible? Which SDKs can I use?
Yes. Apertis exposes the OpenAI-compatible surface — /v1/chat/completions, /v1/responses, /v1/embeddings, /v1/images/generations, /v1/audio/transcriptions — alongside the native Anthropic Messages API at /v1/messages. You can use the OpenAI SDK (Python, JS, Go) by setting base_url to https://api.apertis.ai/v1, the Anthropic SDK by setting the base URL to https://api.apertis.ai, or the first-party Apertis SDK at github.com/apertis-ai/apertis-sdk for richer ergonomics. Streaming, function calling, vision, and tool use all pass through.
How does Apertis pricing work?
Two ways to pay, mix as you like. Subscription tiers are Lite at $12/mo, Pro at $25/mo, Plus at $60/mo, and Max at $180/mo — each with a defined quota for predictable monthly cost. Pay-as-you-go lets you top up credits and pay per million tokens at per-model rates, with no minimum spend. When a subscription quota is exhausted, requests fall back to PAYG credits with a hard cap you control, so a runaway agent loop won't drain your card.
How is Apertis different from OpenRouter / ZenMux / 9router?
Apertis uses paid, direct provider relationships instead of leaning on free quotas a vendor can revoke — reliability does not hinge on a free tier. Apertis survives provider OAuth crackdowns because it is a multi-supplier paid gateway, not an OAuth piggyback. It is operated by Stima AI LLC, a US entity with billing infrastructure built for enterprise procurement including invoices, line-item receipts, and role-based team access. Per-request observability is live at apertis.ai/log to see which model, which agent, which key, in near real time.
Is Apertis safe to use with Claude Code / Cursor / Cline / Roo Code / OpenCode?
Yes — every major coding agent works by setting base_url. Your IDE config doesn't change; only the endpoint does. Setup guides are available for Claude Code, Cursor, OpenCode, Roo Code, and Cline. Because Apertis is multi-supplier, switching from one model to another (e.g. Claude to GPT mid-task) is a single field in the request.
What endpoints does Apertis expose?
Apertis exposes OpenAI-compatible endpoints including /v1/chat/completions, /v1/responses, /v1/embeddings, /v1/images/generations, and /v1/audio/transcriptions. It also exposes the native Anthropic Messages API at /v1/messages. All these endpoints support streaming, function calling, vision, and tool use.
Can I monitor my usage and see which models my agents are using?
Yes. Per-request observability is available at apertis.ai/log where you can see in near real-time which model was used, which agent made the request, and which virtual key was used. This helps with debugging, cost tracking, and enforcing model policy.
What happens when my subscription quota is exhausted?
When a subscription quota is exhausted, requests automatically fall back to PAYG (pay-as-you-go) credits with a hard cap you control. This ensures a runaway agent loop won't drain your card unexpectedly. You can top up PAYG credits at any time with no minimum spend requirement.
Does Apertis offer academic or enterprise discounts?
Apertis has an Academic program for researchers and students. For enterprise needs, custom pricing, SLA, dedicated support, and advanced features are available with custom quotas. Enterprise plans include billing infrastructure built for enterprise procurement with invoices and line-item receipts.