Added support for multiple OpenAI endpoints

CHANGELOG.md

@@ -4,12 +4,16 @@ All notable changes to `jitone-ai` will be documented in this file.
 
 ## v0.0.1-alpha - 2024-09-08
 
-- Alpha Release for Internal Testing
+- Alpha Release for Internal Testing.
 
 ## v0.0.8-beta - 2024-09-09
 
-- Release beta version
+- Release beta version.
 
 ## v0.1.0 - 2024-09-10
 
-- Initial Release
+- Initial Release.
+
+## v0.1.1 - 2024-09-19
+
+- Added support for multiple OpenAI endpoints.

README.md

@@ -7,6 +7,15 @@
 [![GitHub Tests Action Status](https://img.shields.io/github/actions/workflow/status/jiten14/jitone-ai/run-tests.yml?branch=main&label=tests&style=flat-square)](https://github.com/jiten14/jitone-ai/actions?query=workflow%3Arun-tests+branch%3Amain)
 [![Total Downloads](https://img.shields.io/packagist/dt/jiten14/jitone-ai.svg?style=flat-square)](https://packagist.org/packages/jiten14/jitone-ai)
 
+## What's New in Jitone AI v0.1.1
+
+1. **Support for Multiple OpenAI Endpoints**:
+   - The latest OpenAI models, including the `gpt-4` and `gpt-3.5-turbo`, now use the **Chat API endpoint**, improving performance and response flexibility.
+   - The `gpt-3.5-turbo-instruct` model will continue using the **Completion API endpoint**.
+
+2. **Completion Models Now Considered Legacy**:
+   - Please note that the models using the Completion endpoint, such as `gpt-3.5-turbo-instruct`, are now considered **legacy models**.
+   - We highly recommend trying out the latest models for improved results and future-proofing. You can explore them here: [OpenAI Models Documentation](https://platform.openai.com/docs/models).
 
 ## Installation
 
@@ -170,6 +179,7 @@ AIFileUpload::make('image')
 Jitone AI follows semantic versioning:
 
 - **v0.1.0**: Initial release.
+- **v0.1.1**: Added support for multiple OpenAI endpoints.
 
 Please see [CHANGELOG](CHANGELOG.md) for more information on what has changed recently.
 
@@ -178,9 +188,8 @@ Please see [CHANGELOG](CHANGELOG.md) for more information on what has changed re
 Please see [CONTRIBUTING](.github/CONTRIBUTING.md) for details.
 
 ## FAQ
-- **File upload previews not loading after generating Image.**
-
-**Ans.**- Make sure that the APP_URL variable in your .env file matches the domain you're using to access your app from, including the protocol (http or https).
+1. **File upload previews not loading after generating Image.**
+    - **Ans.**- Make sure that the APP_URL variable in your .env file matches the domain you're using to access your app from, including the protocol (http or https).
 
 ## Security Vulnerabilities
 

config/jitone-ai.php

@@ -1,7 +1,7 @@
 <?php
 
 return [
-    'default_model' => 'gpt-3.5-turbo-instruct',
+    'default_model' => 'gpt-4o',
     'default_max_tokens' => 150,
     'default_temperature' => 0.7,
     'default_image_size' => '1024x1024',

src/Services/OpenAIService.php

@@ -12,13 +12,30 @@ public function generateContent(string $prompt, array $options = [])
         $maxTokens = $options['max_tokens'] ?? config('jitone-ai.default_max_tokens');
         $temperature = $options['temperature'] ?? config('jitone-ai.default_temperature');
 
-        $completion = OpenAI::completions()->create([
-            'model' => $model,
-            'prompt' => $prompt,
-            'max_tokens' => $maxTokens,
-            'temperature' => $temperature,
-        ]);
+        if ($model === 'gpt-3.5-turbo-instruct') {
+            $completion = OpenAI::completions()->create([
+                'model' => $model,
+                'prompt' => $prompt,
+                'max_tokens' => $maxTokens,
+                'temperature' => $temperature,
+            ]);
 
-        return $completion['choices'][0]['text'];
+            return $completion['choices'][0]['text'];
+        } else {
+            $result = OpenAI::chat()->create([
+                'model' => $model,
+                'messages' => [
+                    [
+                        'role' => 'user',
+                        'content' => $prompt,
+                    ],
+                ],
+                'max_tokens' => $maxTokens,
+                'temperature' => $temperature,
+            ]);
+
+            return $result->choices[0]->message->content;
+        }
     }
+
 }