简体中文
EnglishWhich AI models and integrations does PromptHub support?
Updated
PromptHub supports testing prompts across OpenAI, Anthropic, Azure, Google, Meta, Bedrock, and Mistral, and it connects to your stack through a REST API, forms, and Zapier. If your team needs to compare model outputs side-by-side or run the latest prompt version from a branch in your own application, those are the two capabilities to check first.
Supported AI models
PromptHub's prompt playground lets you test across different models and compare outputs side-by-side. The models named on the site are:
| Provider | Notes |
|---|---|
| OpenAI | Listed as a test target in the playground |
| Anthropic | Listed as a test target in the playground |
| Azure | Listed as a test target in the playground |
| Listed as a test target in the playground | |
| Meta | Listed as a test target in the playground |
| Bedrock | Listed as a test target in the playground |
| Mistral | Listed as a test target in the playground |
The practical value here is cross-model comparison: you write a prompt once, then run it against multiple providers to see how outputs differ before committing to one. This matters when a prompt behaves well on one model but regresses on another, or when you want a fallback provider.
Integrations
PromptHub exposes three integration paths for getting prompts into your own systems.
API
The documented endpoint runs the latest prompt commit from any branch:
POST https://app.prompthub.us/api/v1/projects/{id}/run?branch=branch%20name
The site describes three return modes — Run Prompt, Return Prompt, and Return All Prompts — so you can choose whether the API executes the prompt or hands back the prompt content for your own runtime to execute. The branch parameter means you can point production at a stable branch and development at a working branch without duplicating prompts.
Forms
Forms are listed as a deployment surface, which fits cases where a non-technical teammate needs to submit inputs to a prompt without touching code.
Zapier
Zapier is listed as a deployment surface, which lets you trigger prompts from events in other tools without building a custom integration.
How to decide whether this fits
- Choose PromptHub for model comparison if you need to evaluate the same prompt across OpenAI, Anthropic, Azure, Google, Meta, Bedrock, and Mistral in one place rather than juggling separate provider consoles.
- Choose the API path if your application needs to fetch or run a specific prompt version at runtime and you want branch-based control over which version is live.
- Choose Forms or Zapier if the consumers of the prompt are non-developers or the trigger lives in another SaaS tool.
- Look elsewhere if you need an integration that isn't API, Forms, or Zapier — the site names only these three, so anything else would require custom work.
What to verify before committing
The site lists a free start and a demo booking, and a pricing page exists, but the input doesn't state which features sit behind which plan. Before rolling this out, confirm two things directly: whether the models you actually use are available in your plan's playground, and whether API rate limits or branch counts are capped. Those are the constraints most likely to affect a production deployment, and they aren't specified in the available material.