AI Namespace

The ai namespace provides AI/LLM service providers.

import { ai } from "nevr-env/plugins";

Providers

Provider	Description
ai.openai()	OpenAI - GPT, DALL-E, Whisper, Embeddings

---

OpenAI

Full-featured OpenAI integration with organization support, Azure OpenAI, and model configuration.

Basic Usage

import { createEnv } from "nevr-env";
import { ai } from "nevr-env/plugins";

const env = createEnv({
  plugins: [ai.openai()],
  runtimeEnv: process.env,
});

// env.OPENAI_API_KEY

Options

Option	Type	Default	Description
organization	boolean	false	Include organization ID
project	boolean	false	Include project ID
model	boolean	false	Include model configuration
defaultModel	string	"gpt-4o"	Default model name
azure	boolean	false	Include Azure OpenAI config
baseUrl	boolean	false	Include custom base URL
embedding	boolean	false	Include embedding model config
parameters	boolean	false	Include default parameters
variableNames	object	-	Custom variable names
extend	object	-	Extend schema with custom fields

Environment Variables by Option

Default (no options)

Variable	Required	Format	Description
OPENAI_API_KEY	✅	sk-*	OpenAI API key

organization: true

Variable	Required	Format	Description
OPENAI_ORG_ID	❌	org-*	Organization ID

project: true

Variable	Required	Description
OPENAI_PROJECT_ID	❌	Project ID for newer OpenAI projects

model: true

Variable	Required	Default	Description
OPENAI_MODEL	❌	gpt-4o	Default model to use

Supported models:

• gpt-4o, gpt-4o-mini
• gpt-4-turbo, gpt-4
• gpt-3.5-turbo
• o1, o1-mini, o1-preview

azure: true

Variable	Required	Default	Description
AZURE_OPENAI_ENDPOINT	✅	-	Azure OpenAI endpoint URL
AZURE_OPENAI_API_VERSION	❌	2024-02-01	API version
AZURE_OPENAI_DEPLOYMENT	❌	-	Deployment name

baseUrl: true

Variable	Required	Description
OPENAI_BASE_URL	❌	Custom base URL (proxies, alternative endpoints)

embedding: true

Variable	Required	Default	Description
OPENAI_EMBEDDING_MODEL	❌	text-embedding-3-small	Embedding model

parameters: true

Variable	Required	Default	Description
OPENAI_MAX_TOKENS	❌	4096	Maximum tokens per request
OPENAI_TEMPERATURE	❌	0.7	Sampling temperature (0-2)

Examples

// With organization and model selection
ai.openai({
  organization: true,
  model: true,
  defaultModel: "gpt-4o",
})

// Azure OpenAI
ai.openai({
  azure: true,
})

// RAG application with embeddings
ai.openai({
  model: true,
  embedding: true,
  parameters: true,
})

// With custom OpenAI-compatible endpoint
ai.openai({
  baseUrl: true,
})

// Extend with assistant configuration
ai.openai({
  extend: {
    OPENAI_ASSISTANT_ID: z.string().startsWith("asst_"),
    OPENAI_VECTOR_STORE_ID: z.string().startsWith("vs_").optional(),
    OPENAI_THREAD_ID: z.string().startsWith("thread_").optional(),
  }
})

Integration Example

// env.ts
import { createEnv } from "nevr-env";
import { ai } from "nevr-env/plugins";

export const env = createEnv({
  plugins: [
    ai.openai({
      organization: true,
      model: true,
    }),
  ],
  runtimeEnv: process.env,
});

// openai.ts
import OpenAI from "openai";
import { env } from "./env";

export const openai = new OpenAI({
  apiKey: env.OPENAI_API_KEY,
  organization: env.OPENAI_ORG_ID,
});

// chat.ts
import { env } from "./env";
import { openai } from "./openai";

export async function chat(prompt: string) {
  const response = await openai.chat.completions.create({
    model: env.OPENAI_MODEL ?? "gpt-4o",
    messages: [{ role: "user", content: prompt }],
  });
  return response.choices[0].message.content;
}

Azure OpenAI Example

// env.ts
import { createEnv } from "nevr-env";
import { ai } from "nevr-env/plugins";

export const env = createEnv({
  plugins: [
    ai.openai({ azure: true }),
  ],
  runtimeEnv: process.env,
});

// azure-openai.ts
import { AzureOpenAI } from "openai";
import { env } from "./env";

export const openai = new AzureOpenAI({
  endpoint: env.AZURE_OPENAI_ENDPOINT,
  apiVersion: env.AZURE_OPENAI_API_VERSION,
  deployment: env.AZURE_OPENAI_DEPLOYMENT,
});