Laravel AI Hub

One interface. Nine AI providers. Full cost & failure observability.

Swap OpenAI, Gemini, Claude, Grok, DeepSeek, Mistral, Groq, Ollama, or OpenRouter via config or a premium visual UI — every call is logged, costed, retried, and (optionally) JSON-repaired automatically. Set provider priority so #1 runs first and failover continues down the chain.

Built by Imran Dev BD · Laravel 10–13 · PHP 8.2+

composer require imrandevbd/laravel-ai-hub

Packagist: imrandevbd/laravel-ai-hub
GitHub: github.com/imranbru99/laravel-ai-hub

Why this package?

Features

Install

composer require imrandevbd/laravel-ai-hub
php artisan vendor:publish --tag=ai-hub-config
php artisan migrate

.env (optional — you can also set keys in the UI)

AI_HUB_PROVIDER=openai
AI_HUB_FAILOVER=true
AI_HUB_MIDDLEWARE=auth   # protect /ai-hub in production

OPENAI_API_KEY=sk-...
GEMINI_API_KEY=...
ANTHROPIC_API_KEY=...
GROK_API_KEY=...
DEEPSEEK_API_KEY=...
MISTRAL_API_KEY=...
GROQ_API_KEY=...
OPENROUTER_API_KEY=...
# OLLAMA_BASE_URL=http://127.0.0.1:11434/v1

Visual studio (/ai-hub)

Open /ai-hub after install.

Tabs

1. Keys & Models — paste API keys, pick models, enable/disable providers, Test connection
2. Priority — reorder who runs first (↑ ↓ / First). Failover walks the list on errors
3. Analytics — 30-day cost, failure %, JSON recovery %, latency p95, cost by provider, top jobs, daily chart

Built with Tailwind CSS + Alpine.js (CDN) — modern dark premium UI.

CLI alternative

php artisan ai-hub:configure
php artisan ai-hub:configure gemini --key=YOUR_KEY --model=gemini-2.0-flash --default
php artisan ai-hub:configure deepseek --key=YOUR_KEY --model=deepseek-chat
php artisan ai-hub:configure --show

Usage

use ImranDevBd\AiHub\Facades\AIHub;

// Shortcuts
$response = AIHub::gemini('gemini-2.0-flash')
    ->prompt('Summarize Laravel queues in 3 bullets')
    ->send();

echo $response->content;
echo $response->costUsd;
echo $response->latencyMs;

// JSON + recovery (great for Gemini)
$data = AIHub::provider('gemini')
    ->prompt('Return JSON: {"title":"","tags":[]}')
    ->recoverJson()
    ->send()
    ->json();

// Uses priority chain (#1 first, then failover)
AIHub::prompt('Hello')->send();

// Lock to one provider
AIHub::provider('openai')->withoutFailover()->prompt('Hello')->send();

// Persist settings for the whole app
AIHub::configure('openai', 'sk-...', 'gpt-4o-mini', makeDefault: true);

// Job tracing (shows in Analytics → Top jobs)
AIHub::gemini()->forJob(MyJob::class)->prompt($prompt)->send();

Streaming & embeddings

foreach (AIHub::openai()->prompt('Write a haiku')->stream() as $chunk) {
    echo $chunk;
}

$vector = AIHub::openai()
    ->model('text-embedding-3-small')
    ->embed('Laravel AI Hub')
    ->first();

Analytics API

AIHub::analytics()->summary(now()->subDays(30));
AIHub::analytics()->costByProvider();
AIHub::analytics()->latencyPercentiles('gemini');
AIHub::analytics()->topJobs(10);
AIHub::analytics()->dailyCost(30);

Priority / failover

Set order in /ai-hub → Priority or config:

// config/ai-hub.php
'priority' => ['gemini', 'openai', 'claude', 'grok', 'deepseek', 'mistral', 'groq', 'ollama', 'openrouter'],
'failover_enabled' => true,

Runtime:

#1 Gemini fails → #2 OpenAI → #3 Claude → …

Providers

AIHub::deepseek()->prompt('Explain recursion')->send();
AIHub::mistral('mistral-large-latest')->prompt($prompt)->send();
AIHub::groq()->prompt('Fast answer')->send();
AIHub::ollama('llama3.2')->prompt('Hi from local')->send();
AIHub::openrouter('anthropic/claude-3.5-sonnet')->prompt($prompt)->send();

Extend later via custom base_url: Azure OpenAI, Cohere, Perplexity, Bedrock proxies, etc.

Package layout

src/
  AIHubManager.php · PendingRequest.php
  Facades/AIHub.php
  Providers/   OpenAI · Gemini · Claude · OpenAICompatible (Grok, DeepSeek, Mistral, Groq, Ollama, OpenRouter)
  Support/     JsonRecovery · RetryHandler · CostCalculator · Analytics · SettingsStore · ProviderCatalog
  Http/Controllers/StudioController.php
  Models/      AiRequestLog · AiHubSetting
  Jobs/        TrackAiUsageJob
  Filament/    AiHubDashboardWidget (optional)
  Console/     ConfigureAiHubCommand
config/ai-hub.php
resources/views/studio.blade.php
database/migrations/

Security

Protect /ai-hub in production:

AI_HUB_MIDDLEWARE=auth

Prefer the visual UI for keys on shared hosts; keep .env out of git.

Changelog

See CHANGELOG.md.

License

MIT © Imran Dev BD

Developed by Imran Dev BD

Any issue? Contact → imrandev.bd/contact