Artisan commands

Command	Description
ai:make-task	Generate a task class
ai:request	Ad-hoc sync or queued request
ai:runs	List recent ai_runs
ai:retry	Re-dispatch failed and stuck runs
ai:budget	Tenant’s spend vs limit
ai:models	Models available from a provider’s API

ai:make-task

php artisan ai:make-task SummarizeTask
php artisan ai:make-task Orders/AnalyzeTask --queued

Creates the class in App\Ai\Tasks (subdirectories via /), with SerializesModelsAi.

Option	Description
--queued	Implement ShouldQueueAi
--force	Overwrite an existing file

ai:request

php artisan ai:request "Summarize the plot of Hamlet"
php artisan ai:request "..." --driver=anthropic --raw
php artisan ai:request "..." --queue

Option	Description
--driver=	Driver name
--modality=text	text or embed
--temperature=	Sent only when given (some models reject it)
--tenant=	Tenant ID
--queue	Dispatch to the queue, print the run id
--json	Print the content only
--raw	Also print usage: tokens, cost

ai:runs

php artisan ai:runs --task=summarize --status=dead --limit=50

Option	Description
--tenant=	Filter by tenant
--task=	Filter by task name
--status=	Filter by status
--limit=20	Rows

ai:retry

Re-dispatches runs with status error or dead: rebuilds the task from task_class/task_args in ai_runs.request and re-queues it into the same row (status goes back to queued).

php artisan ai:retry                      # failures from the last 24h
php artisan ai:retry --since=1h --limit=10
php artisan ai:retry --stuck              # also stuck queued/running runs
php artisan ai:retry --dry-run            # list only, change nothing

Option	Description
--since=24h	Period in hours
--limit=50	Max runs
--stuck	Include stuck runs
--dry-run	Only list what would be retried

Task constructor arguments are stored only with AI_STORE_REQUEST=true — runs recorded without them are listed as skipped. Tasks whose constructor has no required parameters can always be retried.

A failed attempt of a sync call whose fallback driver then answered is not retried — the call already has its result (response.superseded_by points to the row that answered).

ai:budget

php artisan ai:budget tenant-abc
php artisan ai:budget tenant-abc --month=2026-09
php artisan ai:budget tenant-abc --from=2026-09-01 --to=2026-09-15

Argument / option	Description
tenant	Tenant ID, default default
--month=	Month YYYY-MM
--from=, --to=	Date range YYYY-MM-DD; --to defaults to today

ai:models

php artisan ai:models                       # all configured drivers
php artisan ai:models gemini                # one driver
php artisan ai:models openai --filter=gpt-5 # filter by substring
php artisan ai:models anthropic --detail    # token limits, release date, capabilities

The currently configured model is marked with ✓. Groq, Mistral, DeepSeek, xAI, Ollama and OpenRouter are queried via the OpenAI-compatible /v1/models endpoint at each provider’s default URL; set ai.providers.{driver}.url only to override it (e.g. a self-hosted Ollama).

From code — AI::models().