Key facts
| Provider | AWS Bedrock — models from multiple vendors behind AWS-native APIs |
| API style | IAM-authenticated, region-scoped endpoints, including the Converse API |
| Plugsky API | OpenAI-compatible /v1/chat/completions — change the base URL, keep your SDK |
| Models | 30+ models from free to frontier behind one API key |
| Pricing | Flat monthly plans with unlimited fair-use usage; no per-token billing on self-serve |
| Free tier | Free plan with plugsky-micro and plugsky-lite, no card; 14-day full-access trial |
| Deployment | Plugsky cloud, your VPC, on-prem or air-gapped; region choice for residency |
| Live vs roadmap | Chat, streaming, JSON mode, function calling, embeddings, RAG, agents live; audio, images, moderation, files, batch, fine-tuning, assistants, responses coming soon |
TL;DR
- Bedrock suits teams standardized on AWS SDKs, IAM and VPC endpoints.
- Plugsky suits teams standardized on OpenAI-format clients and portable code.
- Plugsky: 30+ models, flat monthly self-serve pricing, no per-token billing.
- Both can be evaluated side by side before committing production traffic.
- Honest trade-off: Bedrock's AWS-native guardrails and vendor breadth remain strengths.
How it works, step by step
- Capture the current Bedrock call patterns your services use (Converse, InvokeModel or streaming).
- Create a Plugsky account and generate an API key on the free plan.
- Write one internal client interface and two adapters — Bedrock and OpenAI-compatible.
- Map model families, then test prompts, tools and JSON mode against both adapters.
- Compare operational signals: latency from your region, error rates and cost shape.
- Route production traffic deliberately and keep rollback to Bedrock one config change away.
Original data
Try it yourself
Open the AWS Bedrock cost calculator →
How the Bedrock API is structured
Bedrock inherits AWS conventions. Requests are signed with IAM credentials, endpoints are region-scoped, and model access is enabled per account and region. The Converse API gives you a consistent message format across model families, which reduces per-vendor branching inside your code.
The cost of those conventions is portability. Application code pulls in AWS SDK dependencies, region selection is constrained by model availability, and local development usually needs credential plumbing before the first request works.
How Plugsky's API is structured
Plugsky speaks the OpenAI dialect. You point your client at api.plugsky.com/v1, keep your SDK and your request shapes, and map model names. That makes the API easy to run locally, easy to test in CI, and easy to swap again later.
The platform adds 30+ models behind one key, a free plan with plugsky-micro and plugsky-lite, and a 14-day full-access trial. Self-serve pricing is flat monthly with unlimited fair-use usage (live pricing), and enterprise deployments can run in your VPC, on-prem or air-gapped. Live endpoints include chat, streaming, JSON mode, function calling, embeddings, RAG and agents; audio, images, moderation, files, batch, fine-tuning, assistants and responses are coming soon.
Choosing between the two shapes
Neither API is universally better; they optimize for different constraints. Pick based on where your code and governance already live, not on a synthetic benchmark.
- If IAM, VPC endpoints and CloudWatch are mandatory, Bedrock's shape is the advantage.
- If your developers already write OpenAI-format code, Plugsky removes integration friction.
- If you want one predictable bill across many workloads, flat monthly plans simplify the maths.
- If you need multi-cloud portability, an OpenAI-compatible surface keeps your options open.
Honest comparison
| Capability | Plugsky | AWS Bedrock | Building in-house |
|---|---|---|---|
| API style | OpenAI-compatible drop-in | IAM-signed, region-scoped AWS APIs (Converse) | You define the schema |
| Auth | API keys with dashboard management | IAM identities, policies and roles | You build identity |
| Model catalogue | 30+ models, one key | Broad vendor list enabled per region | You host each model |
| Billing | Flat monthly, unlimited fair use (see live pricing) | Usage-based, billed through AWS | GPU + ops cost |
| Residency | Region choice, VPC, on-prem, air-gapped | AWS regions and controls | You control the infrastructure |
| Honest gap | Audio, images, batch, fine-tuning coming soon | Deeper AWS governance and vendor breadth | You build it |
Frequently asked questions
Is Bedrock OpenAI-compatible?
Not natively. Bedrock exposes AWS-native APIs, including the Converse API, and some vendors add their own compatibility layers; OpenAI-format clients generally need an adapter.
Is Plugsky a drop-in replacement for the Bedrock API?
For application code written against OpenAI-format clients, yes — change the base URL and model names. Code bound to AWS SDKs needs an adapter layer.
Can I use Plugsky and Bedrock at the same time?
Yes. Keeping both behind one internal interface is a practical way to test quality, cost and availability before consolidating.
Is there a free plan?
Yes — plugsky-micro and plugsky-lite are free with no credit card, and a 14-day full-access trial is available for paid models.
How is Plugsky billed?
Self-serve plans are flat monthly with unlimited fair-use usage; there is no per-token billing on self-serve. See the live pricing page.
Can I keep data in my region?
Plugsky supports region selection and enterprise deployments in your VPC, on-prem or air-gapped; Bedrock provides AWS-region controls.
What does Plugsky not replicate from Bedrock?
AWS-native identity, PrivateLink and Guardrails integration, and Bedrock's full vendor list; Plugsky focuses on OpenAI compatibility, model breadth behind one API, and deployment control.