Premedice API — OpenAI-Compatible Medical AI Endpoint

Q: Which AI models are available through the Premedice API?

The API exposes a single model name, Premedice/Premed-4, which internally routes to one of three tiers based on your subscription: High (fastest responses), Mid (balanced performance), or Low (lowest cost).

Q: What medical tools are included in the API?

The Premedice API includes 15 built-in medical tools: PubMed search, OpenFDA drug information, RxNorm drug lookup, DailyMed labeling, ClinicalTrials.gov search, Semantic Scholar academic search, Europe PMC search, and a create_file tool.

Q: How does billing work on the Premedice API?

The API uses per-token billing. Pricing varies by variant: High (input: $6/1M tokens, output: $30/1M tokens), Mid (input: $5/1M tokens, output: $15/1M tokens), Low (input: $1.20/1M tokens, output: $10/1M tokens).

Q: What are the rate limits for the Premedice API?

Rate limits depend on your subscription plan. Pro plan: 60 requests per minute. Advanced plan: 120 requests per minute.

01 — Authentication

How do I authenticate with the Premedice API?

All requests require a Premedice API key via Bearer token. Generate one from your dashboard.

Generate an API key from your Premedice dashboard. Keys start with pm_live_.

Header

HTTP

Authorization: Bearer pm_live_abc123...

⚠

Requirements: You must have an active paid Premedice subscription and sufficient credits. Free accounts cannot use the API.

Rate Limits

Plan	Requests / min	Monthly Tokens
Pro	60	Included in plan
Advanced	120	Included in plan

02 — Quick Connect

How do I connect to the Premedice API?

Get started in under 2 minutes. Pick your tool.

opencode

Connect from the opencode CLI in two steps:

Step 1 — Run the connect command

Terminal

/connect

Select OpenAI compatible from the provider list, then enter:

Config

Base URL:  https://api.premedice.com/v1
API Key:   pm_live_your_key_here
Model:     Premedice/Premed-4

Step 2 — Or add to your opencode.json directly

Create or edit opencode.json in your project root:

JSON

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "custom": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Medical AI",
      "options": {
        "baseURL": "https://api.premedice.com/v1",
        "apiKey": "pm_live_your_key_here"
      },
      "models": {
        "Premedice/Premed-4": {
          "name": "Premedice 4",
          "limit": {
            "context": 128000,
            "output": 4096
          }
        }
      }
    }
  }
}

✓

That's it. opencode will now route all messages through the API with enhanced research capabilities automatically enabled. No extra configuration needed.

Cursor

Add Premedice as a custom model in Cursor:

1. Open Settings → Models

2. Under "OpenAI API Key", click the gear icon and add:

Config

OpenAI API Key:    pm_live_your_key_here
Override Base URL: https://api.premedice.com/v1

3. Select Premedice/Premed-4 from the model dropdown

ℹ

Cursor tip: Use this API for specialized tasks that benefit from enhanced research capabilities. Keep your regular model for general coding tasks. You can switch models per conversation.

Continue.dev

Add to your ~/.continue/config.yaml:

YAML

models:
  - title: Premedice Medical
    provider: openai
    model: Premedice/Premed-4
    apiBase: https://api.premedice.com/v1
    apiKey: pm_live_your_key_here

VS Code (GitHub Copilot Chat)

Use with Continue.dev extension or any OpenAI-compatible extension:

JSON

{
  "openai.apiKey": "pm_live_your_key_here",
  "openai.baseUrl": "https://api.premedice.com/v1",
  "openai.model": "Premedice/Premed-4"
}

Any OpenAI SDK (Python / Node.js)

Python

import openai

client = openai.OpenAI(
    api_key="pm_live_your_key_here",
    base_url="https://api.premedice.com/v1"
)

response = client.chat.completions.create(
    model="Premedice/Premed-4",
    messages=[{"role": "user", "content": "What is the mechanism of action of metformin?"}]
)
print(response.choices[0].message.content)

JavaScript

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "pm_live_your_key_here",
  baseURL: "https://api.premedice.com/v1",
});

const response = await client.chat.completions.create({
  model: "Premedice/Premed-4",
  messages: [{ role: "user", content: "What is the mechanism of action of metformin?" }],
});
console.log(response.choices[0].message.content);

03 — Chat Completions

How do I create a medical AI chat completion?

The /v1/chat/completions endpoint is identical to OpenAI's API. Send messages, receive medical AI responses.

POST /v1/chat/completions

Request Body

Field	Type	Required	Description
`model`	string	Yes	`Premedice/Premed-4`
`messages`	array	Yes	Array of message objects with `role` and `content`
`stream`	boolean	No	Enable SSE streaming (default: true)
`temperature`	number	No	Sampling temperature 0–2 (default: 0.3)
`max_tokens`	integer	No	Max output tokens (default: 4096)
`top_p`	number	No	Nucleus sampling parameter
`stop`	string\|array	No	Stop sequences
`user`	string	No	Unique user identifier for abuse tracking

Messages Format

JSON

{
  "messages": [
    { "role": "system", "content": "You are a helpful assistant." },
    { "role": "user", "content": "What are the first-line treatments for hypertension?" }
  ]
}

ℹ

Enhanced responses. The API provides enriched, well-sourced answers with automatic research capabilities and 15 built-in medical tools. No extra configuration needed.

Examples

cURL

curl https://api.premedice.com/v1/chat/completions \
  -H "Authorization: Bearer pm_live_abc123..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Premedice/Premed-4",
    "messages": [
      {"role": "user", "content": "What is the CHA2DS2-VASc score for a patient with AF, age 72, hypertension, and diabetes?"}
    ]
  }'

Python

import requests

response = requests.post(
    "https://api.premedice.com/v1/chat/completions",
    headers={
        "Authorization": "Bearer pm_live_abc123...",
        "Content-Type": "application/json",
    },
    json={
        "model": "Premedice/Premed-4",
        "messages": [
            {"role": "user", "content": "What is the CHA2DS2-VASc score for a patient with AF, age 72, hypertension, and diabetes?"}
        ],
    },
)

print(response.json()["choices"][0]["message"]["content"])

JavaScript

const res = await fetch("https://api.premedice.com/v1/chat/completions", {
  method: "POST",
  headers: {
    "Authorization": "Bearer pm_live_abc123...",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    model: "Premedice/Premed-4",
    messages: [
      { role: "user", content: "What is the CHA2DS2-VASc score for a patient with AF, age 72, hypertension, and diabetes?" }
    ],
  }),
});

const data = await res.json();
console.log(data.choices[0].message.content);

OpenAI SDK

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "pm_live_abc123...",
  baseURL: "https://api.premedice.com/v1",
});

const completion = await client.chat.completions.create({
  model: "Premedice/Premed-4",
  messages: [
    { role: "user", content: "What is the CHA2DS2-VASc score for a patient with AF, age 72, hypertension, and diabetes?" }
  ],
});

console.log(completion.choices[0].message.content);

Response

JSON

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1721846400,
  "model": "Premedice/Premed-4",
  "choices": [
    {
      "index": 0,
      "finish_reason": "stop",
      "message": {
        "role": "assistant",
        "content": "For this patient with atrial fibrillation, age 72, hypertension, and diabetes, the CHA\u2082DS\u2082-VASc score is **5**:\n\n- **C**hronic heart failure: 0\n- **H**ypertension: 1\n- **A**ge \u226575: 0\n- **D**iabetes: 1\n- **S**troke/TIA/thromboembolism: 0\n- **V**ascular disease: 0\n- **A**ge 65-74: 1\n- **Sc**ex category (female): 1\n\nA score of 5 indicates high stroke risk. Oral anticoagulation is strongly recommended per current guidelines (ESC 2020, AHA/ACC 2023) [PubMed:32386591].\n\n**Safety note:** Confirm no contraindications to anticoagulation (active bleeding, severe renal/hepatic impairment) before initiating therapy."
      }
    }
  ],
  "usage": {
    "prompt_tokens": 127,
    "completion_tokens": 245,
    "total_tokens": 372
  }
}

Streaming

Streaming is enabled by default. Set "stream": false to disable. Each chunk follows the OpenAI streaming format:

SSE

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","created":1721846400,"model":"Premedice/Premed-4","choices":[{"index":0,"delta":{"role":"assistant","content":"For"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","created":1721846400,"model":"Premedice/Premed-4","choices":[{"index":0,"delta":{"content":" this"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","created":1721846400,"model":"Premedice/Premed-4","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}

data: [DONE]

04 — Models

Which AI models are available through the Premedice API?

The API exposes a single model name. Your subscription plan determines which backend variant handles your requests.

GET /v1/models

Response

JSON

{
  "object": "list",
  "data": [
    {
      "id": "Premedice/Premed-4",
      "object": "model",
      "created": 1721846400,
      "owned_by": "premedice"
    }
  ]
}

Backend Variants

Your subscription plan determines which backend model processes your requests. The API always returns Premedice/Premed-4 as the model name.

Variant	Backend Model	Best For
High	Premed 4 High	Fastest responses, general medical Q&A
Mid	Premed 4 Mid	Balanced depth and speed
Low	Premed 4 Low	Long outputs, budget-conscious usage

ℹ

Automatic fallback. If the primary model fails, the API automatically retries with fallback models. No configuration needed.

05 — Tools

What medical tools are included in the Premedice API?

15 built-in medical research tools are automatically available to the AI during chat completions.

Tool	Description
PubMed	Search biomedical literature via NCBI PubMed
OpenFDA	Drug information, adverse events, and label data from the FDA
RxNorm	Drug names, ingredients, and dosage forms
DailyMed	Official FDA drug labeling and package inserts
ClinicalTrials.gov	Search ongoing and completed clinical trials
Semantic Scholar	Academic paper search with citation context
Europe PMC	Life sciences and biomedical literature
create_file	Generate files from AI responses

ℹ

Automatic tool use. The AI model decides when to invoke tools based on the user's question. No configuration needed — tools are preloaded in every request.

06 — Pricing

How much does the Premedice API cost?

Per-token billing. Costs are debited from your Premedice token balance.

Low

Premed 4 Low

Input $1.20 / 1M

Output $10.00 / 1M

Default

High

Premed 4 High

Input $6.00 / 1M

Output $30.00 / 1M

Mid

Premed 4 Mid

Input $5.00 / 1M

Output $15.00 / 1M

All prices in USD. Costs are debited from your Premedice token balance. Your subscription plan (Pro or Advanced) includes a monthly token allocation.

07 — Errors

What error codes does the Premedice API return?

Standard HTTP status codes with error details.

Status	Type	Cause
`400`	Invalid request	Missing messages, invalid JSON, or request too large
`401`	Unauthorized	Missing, invalid, or revoked API key
`402`	Payment required	Free account or insufficient credits
`429`	Rate limited	Too many requests per minute
`502`	Upstream error	Model provider returned an error. Retry the request.

Error Response Format

JSON

{
  "error": {
    "message": "Invalid or revoked API key",
    "type": "premedice_error",
    "code": 401,
    "request_id": "req_abc123..."
  }
}

08 — FAQ

Frequently Asked Questions

Common questions about the Premedice medical AI API.

What is the Premedice API? ▾

The Premedice API is an OpenAI-compatible medical AI endpoint. It provides chat completions with built-in medical research tools (PubMed, OpenFDA, RxNorm, ClinicalTrials.gov), per-token billing, and works with any OpenAI SDK, Cursor, Continue.dev, or opencode. It is a drop-in replacement for the OpenAI API for healthcare and medical AI use cases.

How do I get an API key? ▾

Generate an API key from your Premedice dashboard at app.premedice.com/dashboard/settings/api-keys. You need an active paid subscription (Pro or Advanced). Free accounts cannot use the API. Keys start with pm_live_.

Which AI models are available? ▾

The API exposes a single model name, Premedice/Premed-4, which internally routes to one of three tiers based on your subscription: High (fastest responses), Mid (balanced performance), or Low (lowest cost). All variants are OpenAI-compatible with automatic fallback.

What medical tools are included? ▾

The API includes 15 built-in medical tools: PubMed search, OpenFDA drug information, RxNorm drug lookup, DailyMed labeling, ClinicalTrials.gov search, Semantic Scholar academic search, Europe PMC search, and a create_file tool. These tools are automatically available to the AI model during chat completions.

How does billing work? ▾

The API uses per-token billing. You are charged based on input and output tokens consumed per request. Pricing varies by variant: High (input: $6/1M tokens, output: $30/1M tokens), Mid (input: $5/1M tokens, output: $15/1M tokens), Low (input: $1.20/1M tokens, output: $10/1M tokens). Costs are debited from your Premedice token balance.

Is the API HIPAA compliant? ▾

The Premedice API is designed with healthcare data privacy in mind. It uses end-to-end encryption, does not store conversation data beyond the request lifecycle, and follows HIPAA-aligned practices. However, you should consult your compliance team to determine if the API meets your specific regulatory requirements.

Can I use it with OpenAI SDK? ▾

Yes. The Premedice API is fully OpenAI-compatible. Use it as a drop-in replacement by setting baseURL to https://api.premedice.com/v1 and using your Premedice API key. It works with the official OpenAI Python and Node.js SDKs, as well as Cursor, Continue.dev, opencode, LiteLLM, and Portkey.

What are the rate limits? ▾

Rate limits depend on your subscription plan. Pro plan: 60 requests per minute. Advanced plan: 120 requests per minute. Monthly token limits are included in your plan.