FAQ
Common questions about using the Wiro API. If you can't find the answer here, contact support.
Sign up at wiro.ai, then create a project at wiro.ai/panel/project. Copy your API key (and secret, if signature-based) and store them securely. You can see them again on the Projects page after you confirm it's you.
Signature-Based is recommended. It uses HMAC-SHA256 so your API secret never leaves your environment. API Key Only is simpler and fine for server-side applications where you control the environment. Use either method only from your own server, never in browser or mobile app code. See Authentication for details.
Yes. Every account has a concurrency limit that controls how many tasks can run at the same time. The limit scales automatically based on your account balance. See Concurrency Limits for the full table.
No. If a task fails (non-zero pexit), you are not charged. Only successfully completed tasks are billed.
Use the POST /Tool/Detail endpoint. The response includes the model's pricing information. If you're using the MCP server, the search_models and get_model_schema tools also return pricing.
Output files are stored on Wiro's CDN and available for a limited time. Download and store any files you need to keep long-term. See Files for details on file management.
Yes. Output URLs returned by Wiro are publicly accessible. Anyone with the URL can access the file. If you need private storage, download the files to your own infrastructure.
Your account is closed right away and its API keys stop working. Its subscriptions stop renewing and end with their current billing period, without a refund. A renewal invoice that is still unpaid is cancelled. Its automatic top-ups stop. 30 days later, or later if something still needs it, such as an active subscription, we permanently delete its data, including your files, tasks, projects and API keys. Until then, write to [email protected] if you want it restored. We keep records the law requires, such as payment and invoice records. See Data Deletion for details.
Connect to the WebSocket at wss://socket.wiro.ai/v1 and register with the socketaccesstoken from your run response. You'll receive events as the task progresses. For simpler integrations, you can poll the Task Detail endpoint.
LLM models return their response in two places: as merged text in debugoutput, and as a structured entry in the outputs array with contenttype: "raw" containing prompt, raw, and separate thinking and answer arrays. For streaming, each task_output WebSocket event carries the latest cumulative snapshot of the output. See LLM & Chat Streaming for details.
Yes. For fileinput parameters, use the {id}Url suffix (e.g., inputImageUrl). For multifileinput, send comma-separated URLs in the field that puts Url before Multiple (e.g., inputDocumentUrlMultiple). For combinefileinput, pass URLs directly in the original parameter. A fileinput or multifileinput with no URL field takes them in the parameter itself. The URL must be a public http(s) address. See Model Parameters.
pexit is the process exit code: "0" means success, any other value means failure. It's the most reliable way to determine if a task succeeded. Always check pexit in the task_postprocess_end event or Task Detail response. See Tasks.
No. The Wiro MCP server is free. You only pay for the model runs you trigger, at standard pricing.
Yes. Install the Wiro AI community node in your n8n instance to access all Wiro models as drag-and-drop nodes in your workflows.
Yes. All models support an optional callbackUrl parameter. When provided, Wiro will POST the task result to your URL when the task completes. It must be a public http(s) address. See Webhook Callback in Model Parameters.
Wiro Agents are autonomous AI assistants that you deploy and manage via the API. Unlike single model runs, agents can connect to third-party services (Twitter, Google Ads, HubSpot, etc.), maintain conversation memory across sessions, and run scheduled tasks automatically. See the Agent Overview for details.
Browse the agent catalog, subscribe to a plan, and deploy via POST /UserAgent/Deploy. After deployment, configure credentials and start the agent. See Agent Overview for the full lifecycle.
Yes. Use the Agent Messaging API to send messages programmatically, receive streaming responses via WebSocket, and get async notifications via webhooks. You can build white-label chat interfaces, webhook-driven pipelines, and multi-agent workflows. See Agent Use Cases.
Agents connect to services like Twitter, Instagram, Google Ads, and HubSpot via OAuth or API keys. Use POST /UserAgentOAuth/OAuthConnect with the provider's credentialkey to initiate OAuth flows, or set API keys via /UserAgent/CredentialUpsert. See Agent Credentials & OAuth.