Get Started¶
Two ways to get an OpenAI-, Anthropic-, Cohere- and Ollama-compatible endpoint in front of Amazon Bedrock. Both give you the same API; pick by where you are today.
Pick Your Path¶
-
Two commands: deploy on AWS with Terraform
Start here — recommended for production.
- 2 Terraform commands → ECS Fargate with HTTPS, auto-scaling, optional WAF
- Hardened container image (
ghcr.io/stdapi-ai/stdapi.ai-community), free under AGPL-3.0, with no subscription - IP-restricted by default — safe to test right away
- Multi-region variants available (EU / US) for data residency
-
One command — Try locally with Docker
Lighter option, for: local development, evaluation, open-source projects.
- One
docker runcommand — no AWS infrastructure to provision - Uses your local AWS credentials (mounted from
~/.aws) to reach Bedrock - No authentication or IP restriction by default — intended for local use only
- Free hardened container image (AGPL-3.0)
- Full API compatibility — same endpoints as production
- One
Looking for reference documentation rather than a quickstart? See Features for what the gateway does, and the API Overview for endpoints, parameters, and SDK usage.
Before you start
Both paths need:
- An AWS account with access to Amazon Bedrock. Create one free if you don't have one.
- AWS credentials configured locally —
aws configureoraws sso login(AWS CLI setup guide).
The AWS Terraform path additionally needs:
- Administrator-level AWS permissions — the module provisions IAM roles and policies, KMS keys, ECS/Fargate, ALB, and networking. A restricted developer profile will fail.
- A sandbox AWS account is strongly recommended for evaluation. Replicate into your target account once you've validated the stack.
- Terraform or OpenTofu >= 1.9.
- Internet access from the tasks, to pull the image from GitHub Container Registry. Without it, mirror the image into your own Amazon ECR.
Not Sure Which to Pick?¶
| You are... | Go with | Why |
|---|---|---|
| A developer evaluating for the first time | Docker (local) | Fastest feedback loop, no cloud resources to tear down. |
| A team lead validating for production use | AWS (Terraform) | Mirrors real deployment. |
| Already running a production workload elsewhere | AWS (Terraform) | Jump straight to the stack you'll operate long-term. |
| Contributing to an open-source project | Docker (local) | AGPL-3.0 container image is free to use and redistribute. |
You can start local and migrate to AWS later — same API, same SDKs, same client-side changes. Point the base URL at your AWS endpoint, add that deployment's API key, and check your model names against what its regions serve.
Pointing a Client at It¶
Both paths are adopted the same way: point your client at the deployment's base URL, give it that deployment's API key, and change the model name only where it differs. Usually it does not — each model is published under the name its provider uses, so a client already asking for claude-opus-5 or gpt-5.6-sol changes nothing but the base URL. Where a name does differ, MODEL_ALIASES maps the name your application already sends onto a model the deployment serves.
What the base URL buys is the catalogue behind it: it spans every provider your regions offer, not one vendor's list. GET /search_models lists it, and the Models page shows the same catalogue now, before you deploy.
After Your First Call¶
- Features — What the gateway supports, substantiated claim by claim
- API Overview — Endpoints, parameters, and SDK usage
- Search Models — Find the right model ID by modality, route, region, or streaming support
- Models — Browse every model with modalities, regions, AWS prices and leaderboard scores
- Use Cases — Open WebUI, n8n, coding assistants, and more
- Configuration — Every environment variable and option
- Troubleshooting — Common first-deployment errors and fixes
- Advanced Deployment — VPC integration, multi-region, cost optimization
- Licensing — AGPL-3.0-or-later terms