Features — AI Gateway for AWS Bedrock¶

stdapi.ai is an AI gateway purpose-built for AWS. It brings full OpenAI API compatibility to AWS Bedrock and AWS AI services, so your team can use their favorite OpenAI-powered applications—ChatGPT-compatible UIs, coding assistants, automation platforms—on AWS infrastructure with zero friction.

Born from deep AWS Solutions Architecture and software engineering expertise—and a genuine passion for AI—stdapi.ai is designed to work seamlessly whether you're an end user, an ops engineer, or a developer: broad API parameter coverage, careful error handling, and deep AWS integration provide a smooth experience out of the box.

How It Works¶

stdapi.ai sits between your applications and AWS AI services, translating OpenAI API calls into native AWS requests. Any tool or SDK that speaks the OpenAI protocol connects instantly—no plugins, no custom integrations.

%%{init: {'flowchart': {'htmlLabels': true}} }%%
flowchart LR
  openwebui["<img src='../styles/logo_openwebui.svg' style='height:48px;width:auto;vertical-align:middle;' /> Open WebUI"] --> stdapi["<img src='../styles/logo.svg' style='height:64px;width:auto;vertical-align:middle;' /> stdapi.ai"]
  n8n["<img src='../styles/logo_n8n.svg' style='height:48px;width:auto;vertical-align:middle;' /> n8n"] --> stdapi
  ide["<img src='../styles/logo_vscode.svg' style='height:48px;width:auto;vertical-align:middle;' /> IDE + AI Assistant"] --> stdapi
  app["<img src='../styles/logo_openai.svg' style='height:48px;width:auto;vertical-align:middle;' /> Any OpenAI App"] --> stdapi
  stdapi --> bedrock["<img src='../styles/logo_amazon_bedrock.svg' style='height:48px;width:auto;vertical-align:middle;' /> AWS Bedrock"]
  bedrock --> claude["<img src='../styles/logo_anthropic_claude.svg' style='height:36px;width:auto;vertical-align:middle;' /> Claude"]
  bedrock --> qwen["<img src='../styles/logo_qwen.svg' style='height:36px;width:auto;vertical-align:middle;' /> Qwen"]
  bedrock --> mistral["<img src='../styles/logo_mistralai.svg' style='height:36px;width:auto;vertical-align:middle;' /> Mistral"]
  bedrock --> stability["<img src='../styles/logo_stabilityai.svg' style='height:36px;width:auto;vertical-align:middle;' /> Stability AI"]
  bedrock --> more["✨ and more..."]
  stdapi --> transcribe["<img src='../styles/logo_amazon_transcribe.svg' style='height:48px;width:auto;vertical-align:middle;' /> AWS Transcribe"]
  stdapi --> polly["<img src='../styles/logo_amazon_polly.svg' style='height:48px;width:auto;vertical-align:middle;' /> AWS Polly"]
  stdapi --> s3["<img src='../styles/logo_amazon_s3.svg' style='height:48px;width:auto;vertical-align:middle;' /> Amazon S3"]

OpenAI API Compatibility¶

stdapi.ai provides broad OpenAI API compatibility, covering routes and parameters far beyond basic chat. Your existing applications, SDKs, and tools—from Open WebUI to n8n to coding assistants—work immediately.

Supported API Routes¶

Unified Multi-Modal API¶

Access text, image, audio, and video capabilities through a single, consistent API interface:

• Text — Chat completions, embeddings, and translation across 80+ models
• Images — Generation, editing, and variations via Stable Diffusion, Nova Canvas, and more
• Audio — Speech synthesis (Polly), transcription with speaker diarization (Transcribe), and translation
• Video & Documents — Multi-modal inputs in chat completions for models that support them

Broad Parameter Coverage¶

Unlike minimal adapters, stdapi.ai works to map as many OpenAI API parameters as possible to their Bedrock equivalents—across all supported routes, not just chat completions:

• Generation controls — temperature, max_tokens, streaming, and many more specialized parameters
• Tool/function calling with OpenAI-compatible schema, including parallel tool calls
• Streaming via Server-Sent Events (SSE) with token usage reporting
• All message roles — System, developer, user, assistant, and tool
• Image & audio parameters — Size, quality, format, voice, speed, and other route-specific options
• Model-specific features — Support for capabilities unique to specific models (e.g., reasoning effort, prompt caching, system tools)
• Extra parameters — Pass additional model-specific or route-specific parameters beyond the standard OpenAI API via the extra_body field

Purpose-Built for AWS¶

stdapi.ai is engineered specifically for AWS, unlocking advanced Bedrock features and native AI services that generic gateways cannot provide.

Multi-Region Bedrock Access¶

• Configure multiple AWS regions to access the widest selection of models and maximize availability
• Automatic cross-region inference profile selection — stdapi.ai intelligently selects the best inference profile or falls back to direct model invocation
• Region-aware optimization — Models are routed to the optimal region based on availability and your configuration

Advanced Bedrock Features¶

stdapi.ai exposes Bedrock-specific capabilities through the familiar OpenAI API:

AWS AI Services Integration¶

Beyond Bedrock, stdapi.ai integrates natively with AWS AI services—all accessible through OpenAI-compatible endpoints:

• Amazon Polly — High-quality text-to-speech with multiple voices and languages
• Amazon Transcribe — Speech-to-text with speaker diarization support
• Amazon Translate — Language translation for audio translation workflows
• Amazon Comprehend — Automatic language detection for routing

S3 Integration¶

• S3 bucket support for file storage in image and audio operations
• Regional S3 buckets for multi-region deployments
• S3 Transfer Acceleration for faster file access via generated HTTP links

Compliance & Data Sovereignty¶

stdapi.ai gives you full control over where your data is processed, making it straightforward to meet regulatory requirements.

• Region restrictions — Configure exactly which AWS regions are allowed for inference, matching your GDPR, HIPAA, or FedRAMP requirements
• Cross-region inference profile filtering — Easily restrict cross-region profiles to only compliant regions
• Data stays in your AWS account — All inference runs within your own account; data is never shared with model providers or used for training
• No external data transmission — stdapi.ai processes requests locally and communicates only with AWS services

Security¶

Security is built into every layer of stdapi.ai:

Observability & Debugging¶

Monitor, debug, and audit your AI gateway with built-in tooling:

• OpenTelemetry integration — Export traces and metrics to AWS X-Ray, Datadog, or any OTLP-compatible backend
• Request/response logging — Optional detailed logging of full request and response payloads for debugging
• Token usage tracking — Accurate token consumption reporting in API responses
• Swagger & ReDoc interfaces — Interactive API documentation served directly by the application
• Configurable log levels — Fine-grained control over logging verbosity
• Client IP logging — Optional client IP tracking for audit trails

Quality of Life¶

Features that make day-to-day operations smoother:

• Model aliases & overrides — Map custom model names to specific Bedrock model IDs for simplified client configuration
• Model auto-detection — Automatically discovers available Bedrock models in your configured regions
• Model list caching — Cached model listings for fast responses without repeated AWS API calls
• Token usage reporting — Consistent usage statistics across all endpoints
• Zero-configuration startup — Works out of the box with automatic region and model detection

Deployment¶

stdapi.ai offers flexible deployment options for every stage:

• Comprehensive documentation — Detailed Getting Started guide, Configuration Reference, and Use Case walkthroughs
• High-performance runtime — Powered by Granian, a fast Python ASGI server, with configurable workers and threads

Feature Summary¶

A quick-reference checklist to find what you need at a glance:

API & Compatibility¶

• OpenAI Chat Completions API (/v1/chat/completions)
• OpenAI Embeddings API (/v1/embeddings)
• OpenAI Images API (generations, edits, variations)
• OpenAI Audio API (speech, transcriptions, translations)
• OpenAI Models API (/v1/models)
• Streaming (Server-Sent Events)
• Tool / function calling
• Multi-modal inputs (text, image, audio, video, documents)
• Broad parameter mapping (all routes)
• Model-specific features & extra parameters

AWS Integration¶

• Multi-region Bedrock access
• Automatic cross-region inference profile selection
• Prompt caching
• Reasoning modes (extended thinking)
• Bedrock Guardrails
• Service tiers
• Application inference profiles
• Prompt routers
• Amazon Polly (text-to-speech)
• Amazon Transcribe (speech-to-text with diarization)
• Amazon Translate
• S3 integration with Transfer Acceleration

Security & Compliance¶

• API keys in SSM Parameter Store / Secrets Manager
• Region-based data sovereignty controls
• CORS, trusted hosts, proxy headers
• CSRF protection
• Hardened Docker image (commercial)
• Data never leaves your AWS account

Operations¶

• OpenTelemetry integration
• Request/response detail logging
• Swagger & ReDoc API docs
• Model aliases & overrides
• Model auto-detection & caching
• Token usage tracking
• Zero-configuration startup
• Terraform module for production (ECS)
• Community Docker image for development

Ready to Get Started?¶