README

Official PHP / Laravel SDK for the Rodium AI API — unified access to AI models (OpenAI, Anthropic, Google, DeepSeek, MiniMax…) with RODI credit billing and Mobile Money top-ups.

OpenAI-compatible REST API: same endpoints and payloads as documented at rodiumai.io/docs.

Links

Resource	URL
Packagist (Composer install)	packagist.org/packages/rodiumai/laravel-sdk
Source code	github.com/lecodeur228/rodiumai-laravel-sdk
API documentation	rodiumai.io/docs
Dashboard & API keys	rodiumai.io/dashboard
Model catalogue	rodiumai.io/models

Official documentation
Requirements
Installation
Configuration
Quick start
Dynamic models (API)
Multilingual
Chat parameters
Streaming (SSE)
Listing models
Error handling
SDK reference
Testing & development
Contributing
License

Official documentation

Topic	Rodium AI link
Quickstart	rodiumai.io/docs
API overview	docs/api/overview
Chat completions	docs/api/chat-completions
Streaming SSE	docs/api/streaming
Models	docs/api/models · Catalogue
HTTP errors	docs/api/errors

Detailed SDK ↔ API mapping: docs/api-alignment.md.

Requirements

PHP 8.1+ with the json extension
Laravel 10, 11, 12, or 13 (optional — the client works in plain PHP)
Laravel 12+: PHP 8.2+ → use ^0.1.1 minimum
Rodium AI account + API key: dashboard

Installation

Install from Packagist:

composer require rodiumai/laravel-sdk

For Laravel 12 and 13 (PHP 8.2+):

composer require rodiumai/laravel-sdk:^0.1.1

After installation

Publish config (optional but recommended):

php artisan vendor:publish --tag=rodiumai-config

Add your API key to .env (see Configuration).
The ServiceProvider and RodiumAI Facade are auto-discovered — nothing to register in bootstrap/providers.php.

Plain PHP (no Laravel)

Same Composer command. Then instantiate RodiumAI\RodiumAIClient directly (see Quick start).

Configuration

.env file:

RODIUMAI_API_KEY=rd_sk_your_secret_key
RODIUMAI_DEFAULT_MODEL=openai/gpt-4o
RODIUMAI_TIMEOUT=30
# RODIUMAI_LOCALE=fr   # optional — SDK error hints only

The SDK always uses https://api.rodiumai.io/v1 (not configurable).

Never commit .env or API keys to the repository.

Quick start

Laravel (Facade)

use RodiumAI\Facades\RodiumAI;

$catalogue = RodiumAI::models();
$modelId = $catalogue->chatModels()[0]->id;

$response = RodiumAI::model($modelId)
    ->temperature(0.7)
    ->maxTokens(300)
    ->chat('Explain Rodium AI in two sentences.');

echo $response->content;

Plain PHP

use RodiumAI\RodiumAIClient;

$client = new RodiumAIClient(apiKey: getenv('RODIUMAI_API_KEY'));

$catalogue = $client->models();
$modelId = $catalogue->chatModels()[0]->id;

$response = $client->model($modelId)->chat('Hello!');

echo $response->content;

Equivalent to the cURL / OpenAI SDK quickstart: POST https://api.rodiumai.io/v1/chat/completions with Authorization: Bearer {RODIUMAI_API_KEY}.

Dynamic models (API)

Models are fetched live from GET /v1/models — always up to date with the platform catalogue:

$catalogue = RodiumAI::models();

$catalogue->ids();                    // all model IDs
$catalogue->chatModels();             // text/chat-capable only
$catalogue->providerPrefixes();       // ['anthropic', 'google', 'openai', …]
$catalogue->byProvider('anthropic');  // filter by provider prefix
$catalogue->findById('openai/gpt-4o'); // ?ModelInfo

foreach ($catalogue->chatModels() as $info) {
    echo $info->id . ' — ' . $info->contextWindow . PHP_EOL;
}

Pass any id from the API to ->model('openai/gpt-4o').

Multilingual (optional)

Language is never required. By default, SDK error hints are in English and the AI follows the user's message language.

Optional — localized error hints via config:

RODIUMAI_LOCALE=fr

Optional — force AI responses in a specific language:

RodiumAI::language('fr')  // only when you need it
    ->model('openai/gpt-4o')
    ->chat('Bonjour !');

Supported: en (default), fr, es.

Chat parameters

Aligned with chat-completions:

API parameter	SDK
`model`	`->model('openai/gpt-4o')` / `$options['model']` / IDs from `models()`
`messages`	Array of `{role, content}` or `string` (→ `user` message)
`max_tokens`	`->maxTokens()` / `$options['max_tokens']`
`temperature` (0–2)	`->temperature()` / `$options['temperature']`
`top_p` (0–1)	`->topP()` / `$options['top_p']`
`stop`	`$options['stop']`
`stream`	Set automatically by `->stream()`

use RodiumAI\Data\ChatMessage;

$messages = [
    ChatMessage::system('You are a Laravel assistant.'),
    ChatMessage::user('What is a Service Provider?'),
];

$response = RodiumAI::model('openai/gpt-4o')
    ->temperature(0.5)
    ->topP(0.9)
    ->maxTokens(500)
    ->chat($messages);

Streaming (SSE)

Follows docs/api/streaming: data: … lines, end with data: [DONE].

foreach (RodiumAI::model('openai/gpt-4o')->stream('Tell a short story.') as $delta) {
    echo $delta;
}

Laravel — `StreamedResponse`

use Illuminate\Http\Request;
use Symfony\Component\HttpFoundation\StreamedResponse;
use RodiumAI\Facades\RodiumAI;

public function streamChat(Request $request): StreamedResponse
{
    return response()->stream(function () use ($request) {
        foreach (RodiumAI::stream($request->string('message')) as $delta) {
            echo 'data: ' . json_encode(['delta' => $delta]) . "\n\n";
            ob_flush();
            flush();
        }
        echo "data: [DONE]\n\n";
    }, 200, [
        'Content-Type' => 'text/event-stream',
        'Cache-Control' => 'no-cache',
        'X-Accel-Buffering' => 'no',
    ]);
}

Listing models

GET /v1/models — RODI pricing and metadata:

$models = RodiumAI::models();

foreach ($models->ids() as $id) {
    echo $id . PHP_EOL;
}

$anthropic = $models->byProvider('anthropic');

Error handling

See docs/api/errors.

HTTP	SDK exception	Suggested action
401	`UnauthorizedException`	Check `RODIUMAI_API_KEY`
402	`InsufficientCreditsException`	Top up RODI credits (dashboard)
429	`RateLimitException`	Exponential backoff, then retry
422	`ValidationException`	Fix `model` / `messages`
500+	`RodiumAIException`	Retry once, then contact support

use RodiumAI\Exceptions\InsufficientCreditsException;
use RodiumAI\Exceptions\RodiumAIException;
use RodiumAI\Facades\RodiumAI;

try {
    $response = RodiumAI::chat('Test');
} catch (InsufficientCreditsException $e) {
    logger()->warning('Insufficient RODI', ['body' => $e->responseBody()]);
} catch (RodiumAIException $e) {
    logger()->error('Rodium AI', ['code' => $e->getCode(), 'body' => $e->responseBody()]);
}

SDK reference

Method	Returns	Description
`chat($messages, $options = [])`	`ChatResponse`	Non-streaming completion
`stream($messages, $options = [])`	`Generator<string>`	SSE text deltas
`models()`	`ModelCollection`	Catalogue + pricing
`model($id)`	`static`	Fluent: model
`temperature($f)`	`static`	Fluent: 0–2
`topP($f)`	`static`	Fluent: 0–1
`maxTokens($n)`	`static`	Fluent: token limit
`systemPrompt($s)`	`static`	Fluent: system message

DTOs: ChatResponse, ChatMessage, ModelCollection.

Versions

Version	Notes
v0.1.1	Laravel 12 support (`illuminate/support` ^12)
v0.1.0	Initial release: chat, stream, models, enums, Facade

Full history: CHANGELOG.md.

Testing & development

composer install
composer test                 # PHPUnit (mocked HTTP)
export RODIUMAI_API_KEY="…"
php bin/smoke-test.php        # Live API walkthrough in the terminal

Contributing

See CONTRIBUTING.md and docs/architecture.md.

Maintainers: docs/PUBLISHING.md (tags, Packagist).

License

MIT — see LICENSE.

rodiumai / laravel-sdk

Maintainers

Package info

Statistics

Security

README

Links

Table of contents

Official documentation

Requirements

Installation

After installation

Plain PHP (no Laravel)

Configuration

Quick start

Laravel (Facade)

Plain PHP

Dynamic models (API)

Multilingual (optional)

Chat parameters

Streaming (SSE)

Laravel — `StreamedResponse`

Listing models

Error handling

SDK reference

Versions

Testing & development

Contributing

License

rodiumai / laravel-sdk

Maintainers

Package info

Statistics

Security

README

Links

Table of contents

Official documentation

Requirements

Installation

After installation

Plain PHP (no Laravel)

Configuration

Quick start

Laravel (Facade)

Plain PHP

Dynamic models (API)

Multilingual (optional)

Chat parameters

Streaming (SSE)

Laravel — StreamedResponse

Listing models

Error handling

SDK reference

Versions

Testing & development

Contributing

License

Laravel — `StreamedResponse`