Prompts API

Version-Controlled Prompt Configs

Store prompt configurations with full versioning support. Switch model providers, A/B test prompts, and integrate with any application. Your code stays clean, prompts stay flexible.

Get API Access View API Docs
1

Create Prompt Config

Define prompt template, model, and parameters via Dashboard or API

2

Integrate via API

Call the Prompts API from your application with version number

3

Test New Versions

Create new versions and test in parallel without affecting production

4

Switch Instantly

Update version number in API call to deploy new prompt configuration

Enterprise-Grade Prompt Management

Everything developers need for production AI applications. Versioning, provider switching, and zero-downtime updates.

Full Version Control

Maintain 10+ versions of prompt configurations. Test new versions without breaking production. Roll back instantly if needed.

  • 10+ versions per prompt configuration
  • Version-specific API calls
  • Production-safe testing
  • Instant rollback to any version
  • Audit trail for all changes

Model Provider Switching

Switch between OpenAI, Anthropic, Google Gemini, and more without changing your application code. Just update the API configuration.

  • OpenAI (GPT-4, GPT-3.5)
  • Anthropic (Claude 3.5 Sonnet, Opus)
  • Google (Gemini Pro, Ultra)
  • Meta (Llama models)
  • Switch providers in seconds

Developer-First Integration

RESTful API with TypeScript SDK available now. Python and C# SDKs coming soon. Complete documentation and code examples.

  • RESTful API with API key authentication
  • TypeScript/JavaScript SDK (available now)
  • Python SDK (coming soon)
  • C# SDK (coming soon)
  • Comprehensive API docs
  • Code examples and tutorials

Persistent Configuration Storage

Store prompt templates, system instructions, and parameters centrally. Your application logic stays clean and unchanged.

  • Centralized prompt storage
  • System instructions management
  • Parameter configuration
  • Template variables
  • Application logic unchanged

Zero-Downtime Updates

Update prompts without redeploying your application. Test new versions in parallel, then switch production traffic instantly.

  • Hot-swap prompt versions
  • No application redeployment
  • A/B testing support
  • Gradual rollout capabilities
  • Zero downtime guaranteed

Enterprise Security

API keys with granular permissions, rate limiting, and audit logs. Enterprise-grade security with encryption at rest and in transit.

  • API key management
  • Granular permission controls
  • Rate limiting and quotas
  • Detailed audit logs
  • Enterprise security standards

Prompts API Methods

Three simple endpoints to run your prompts: non-streaming for complete responses, or streaming for real-time output.

Non-Streaming

POST /prompt-runs/run-sync

Get the complete response in a single API call. Best for batch processing and when you don't need real-time streaming.

HTTP/2 Streaming

POST /prompt-runs/async

Stream responses using HTTP/2 protocol. Get tokens as they're generated for a responsive user experience.

SSE Streaming

POST /prompt-runs/run-sse

Server-Sent Events streaming for real-time responses. Compatible with EventSource API in browsers.

Full API Reference

For complete details on request/response formats, parameters, error codes, and advanced usage, visit our comprehensive API documentation.

View Complete API Documentation

Simple REST API Integration

Just a simple HTTP POST request. Use curl, any HTTP library, or our optional SDKs—your choice.

  • Version-Specific Calls
    Reference exact version numbers for reproducible results
  • Variable Injection
    Pass replacements at runtime to populate prompt templates
  • No Redeployment
    Update prompts without touching application code
# Non-streaming: Get complete response in one call
curl -X POST 'https://api.hyperleapai.com/prompt-runs/run-sync' \
  -H 'Content-Type: application/json' \
  -H 'x-hl-api-key: YOUR_API_KEY' \
  -d '{
    "promptId": "your-prompt-id",
    "promptVersionId": "version-id",
    "replacements": {
      "topic": "AI in Healthcare",
      "tone": "professional"
    }
  }'

# HTTP/2 Streaming: Get response as it's generated
curl -X POST 'https://api.hyperleapai.com/prompt-runs/async' \
  -H 'Content-Type: application/json' \
  -H 'x-hl-api-key: YOUR_API_KEY' \
  -d '{
    "promptId": "your-prompt-id",
    "promptVersionId": "version-id",
    "replacements": {"topic": "AI"}
  }'

# SSE Streaming: Server-Sent Events for real-time streaming
curl -X POST 'https://api.hyperleapai.com/prompt-runs/run-sse' \
  -H 'Content-Type: application/json' \
  -H 'x-hl-api-key: YOUR_API_KEY' \
  -d '{
    "promptId": "your-prompt-id",
    "promptVersionId": "version-id",
    "replacements": {"topic": "AI"}
  }'

Built for Production

From startups to enterprises, Prompts API powers mission-critical applications

Content Generation

Blog posts, emails, social media

10+ versions for different tones

Data Extraction

Parse invoices, contracts, receipts

A/B test extraction accuracy

Classification & Tagging

Sentiment analysis, topic detection

Switch models without code changes

Text Transformation

Summarization, translation, rewriting

Roll back bad versions instantly

Why Developers Choose Prompts API

Separation of Concerns

Keep prompts separate from application logic. Marketing can tweak copy, developers focus on features.

Production Safety

Test new prompts in staging, roll out gradually, roll back instantly if issues arise. No code changes needed.

Provider Flexibility

Switch between GPT-4, Claude, Gemini without touching code. Optimize for cost, quality, or speed.

Frequently Asked Questions

Everything you need to know

Why use an API for prompts?

Decoupling prompts from code allows you to iterate on AI logic without redeploying your app. It also enables non-developers to manage prompts.

Does it add latency?

Our API adds negligible latency (<20ms) as it simply proxies the request to the underlying model provider with your configured template.

Can I A/B test prompts?

Yes, you can deploy different versions to different user segments by calling specific version IDs in your application logic.

Which models are supported?

We support all major models including GPT-4o, Claude 3.5 Sonnet, Gemini 1.5 Pro, Llama 3, and Mistral.

Is it secure?

Yes, we use enterprise-grade encryption. API keys have granular scopes, and we do not store your input/output data for training purposes.

Still have questions?

Contact our support team

Start Building with Prompts API

Version-controlled prompts, model switching, zero-downtime updates. Everything you need for production AI. Start with our free plan.

Get API Access Read Documentation