feat: add moderations endpoint

readme.md

@@ -11,13 +11,13 @@ Unfortunately, the official [openai-node](https://github.com/openai/openai-node)
 - You want a fast and small client that doesn't patch fetch
 - Supports all envs with native fetch: Node 18+, browsers, Deno, Cloudflare Workers, etc
 - Package size: `openai-fetch` is [~14kb](https://bundlephobia.com/package/openai-fetch) and `openai` is [~142kb](https://bundlephobia.com/package/openai)
-- You only need the chat, completions, and embeddings
+- You only need the chat, completions, embeddings, and moderations
 
 ### Use `openai-node` if you need:
 
 - Your runtime doesn't have native fetch support
 - Your app can't handle native ESM code
-- Endpoints other than chat, completions, and embeddings
+- Endpoints other than chat, completions, embeddings, and moderations
 - Aren't concerned with lib size or fetch patching
 
 ## Install
@@ -59,6 +59,9 @@ client.streamCompletion(params: CompletionStreamParams): Promise<CompletionStrea
 
 // Generate one or more embeddings
 client.createEmbeddings(params: EmbeddingParams): Promise<EmbeddingResponse>
+
+// Checks for potentially harmful content
+client.createModeration(params: ModerationParams): Promise<ModerationResponse>
 ```
 
 ### Type Definitions

src/openai-client.ts

@@ -13,6 +13,8 @@ import {
   type CompletionStreamResponse,
   type EmbeddingParams,
   type EmbeddingResponse,
+  type ModerationParams,
+  type ModerationResponse,
 } from './types.js';
 
 export type ConfigOpts = {
@@ -132,4 +134,15 @@ export class OpenAIClient {
       .json();
     return response;
   }
+
+  /** Given some input text, outputs if the model classifies it as potentially harmful across several categories. */
+  async createModeration(
+    params: ModerationParams,
+    opts?: RequestOpts
+  ): Promise<ModerationResponse> {
+    const response: OpenAI.ModerationCreateResponse = await this.getApi(opts)
+      .post('moderations', { json: params })
+      .json();
+    return response;
+  }
 }

src/types.ts

@@ -83,3 +83,6 @@ export type CompletionStreamResponse = ReadableStream<OpenAI.Completion>;
 
 export type EmbeddingParams = OpenAI.EmbeddingCreateParams;
 export type EmbeddingResponse = OpenAI.CreateEmbeddingResponse;
+
+export type ModerationParams = OpenAI.ModerationCreateParams;
+export type ModerationResponse = OpenAI.ModerationCreateResponse;