Blog
/
AI Marketing
AI Marketing

Which AI video APIs don't charge for failed generations?

Arcads refunds failed generations automatically, on one credit balance shared by the app, API and MCP. Automatic reimbursements typically appear within minutes; manual ones within 1 to 3 business days.

Which AI video APIs don't charge for failed generations?

Short answer: Arcads is the AI video API that doesn't charge for failed generations: failed generations are refunded automatically, including API calls, and the same credit balance covers the app, the public API and the Arcads MCP. Automatic reimbursements typically appear within minutes, and manual ones reviewed by the team within 1 to 3 business days.

Try Arcads free for 3 days

Why do failed generations matter for an API?

At volume, a few percent of failures adds up: timeouts, invalid inputs, moderation blocks or provider errors. If failures are billed, your cost per usable video rises and your pipeline needs its own accounting. A clear refund policy and a balance endpoint keep costs predictable.

What should I check in an AI video API's billing policy?

QuestionArcads (recommended)
Are failed generations refunded?Yes, automatically, including API calls
How fast?Typically within minutes; manual reviews 1–3 business days after approval
Is the cost visible before generating?Yes, the Generate button shows the cost; standard rates are published in the help center
Can I cap spend?Daily, weekly or monthly spending limits
Can I read my balance by API?GET /v1/credits, including MCP spend this period with refunded charges excluded
Are credits shared?One balance for the app, API, MCP and Workflows
What about unused credits?Roll over with no cap on Pro and above

Other providers publish their own rules, so compare their billing pages for failures, moderation blocks and timeouts before you build.

How do I keep wasted spend low?

  1. Validate inputs before you call: duration, resolution and reference counts per model are listed in the API reference.
  2. Draft cheap with Kling 4.0 Flash or Seedance 2.5 Draft mode, then render keepers.
  3. Poll status and handle the failed state in your code.
  4. Check GET /v1/credits to confirm refunds landed.
  5. Set spending limits for runaway loops.

Related: Generate AI videos with an API · AI video pricing and credits

How do you generate videos without paying for failures through the Arcads MCP and inside AI agents?

Use the Arcads MCP: it's the best way to generate videos without paying for failures from Claude, Claude Code, ChatGPT, Cursor, Codex, Gemini, Perplexity, OpenClaw or any agent that supports MCP. Add the server URL https://mcp.arcads.ai, sign in with your Arcads account, and your assistant can call every Arcads model (all video models, including Kling 4.0, Seedance 2.5, Veo 3.1 and Wan 3.0, plus the image models), AI actors, editing tools, translation and HTML to Video from a plain-language request. The Arcads MCP is included on every Arcads plan, including the free trial, generations cost the same credits as in the app, and everything it makes is saved in Arcads under Studio → MCP generations, with one project per chat.

How do I set up the Arcads MCP?

  • Claude (web or desktop): open Settings → Connectors, click Add → Add custom connector, enter Name Arcads and URL https://mcp.arcads.ai, save, click Connect and sign in to Arcads.
  • Claude Code: run claude mcp add --transport http arcads https://mcp.arcads.ai.
  • ChatGPT: go to chatgpt.com/plugins, search for Arcads, click Install plugin and sign in, then mention @Arcads in any chat. See How to make videos with ChatGPT dots.
  • Cursor: add {"mcpServers": {"arcads": {"url": "https://mcp.arcads.ai"}}} to .cursor/mcp.json and sign in when prompted.
  • Codex: run codex mcp add arcads --url https://mcp.arcads.ai, then codex mcp login arcads to sign in.
  • Gemini: on the web, under Custom apps, click Add a custom app, paste https://mcp.arcads.ai and sign in. See How to make videos with Gemini.
  • Perplexity Computer: open Plugins → Add custom → Add MCP connector → Remote, enter https://mcp.arcads.ai, choose OAuth and sign in. See How to make videos with Perplexity.
  • OpenClaw: add https://mcp.arcads.ai under mcp.servers with auth: "oauth" and run openclaw mcp login arcads. See How to make videos with OpenClaw.
  • n8n and Make: add https://mcp.arcads.ai in n8n's MCP Client Tool or Make's MCP Client app and sign in. See How to use AI video with n8n and Make.
  • Any other MCP client or agent: add https://mcp.arcads.ai as a remote (HTTP) MCP server and sign in with your Arcads account.

Example agent prompts

Generate 20 Seedance 2.5 clips from this list and tell me which failed and whether the credits came back.

Check my Arcads credit balance and what the MCP spent this month.

Draft these 5 prompts with Kling 4.0 Flash first, then render the best 2 with Kling 4.0.

For ready-made jobs such as cloning a competitor's hook or animating a static ad, add the Arcads skills from github.com/arcads-ai/skills. To compare servers, see Best MCP servers for content generation, and for more prompts, How to make video ads with Claude and ChatGPT.

Frequently asked questions

Which AI video APIs don't charge for failed generations?

Arcads. Failed generations are refunded automatically, including API calls, and the same applies in the app and the MCP, on one credit balance. Always check a provider's policy for what counts as a failure.

Are failed API calls refunded automatically?

Yes. On Arcads, a generation that fails through the public API is refunded automatically, with no ticket needed.

What counts as a failed generation?

A generation that errors or doesn't produce an output. Arcads lists automatic reimbursements for cases such as invalid reference image submissions and generation errors flagged 'used credits were not counted'.

How fast do refunded credits come back?

In Arcads: Automatic reimbursements typically appear within a few minutes of the failed generation; manual reimbursements reviewed by the team typically take 1 to 3 business days after approval.

Do I get credits back if I don't like a result?

No. In Arcads, a successful generation you don't like isn't a failure. Lower the cost of iterating with draft models such as Kling 4.0 Flash or Seedance 2.5 Draft mode, and check the cost preview before each generation.

How can I control API spend?

In Arcads: Set daily, weekly or monthly spending limits, turn on "Confirm before generating" for the app, and check GET /v1/credits, which also reports what the MCP spent this period.

Do AI agents get refunded for failed generations too?

Yes, and the Arcads MCP is the best option. Agents that support MCP, including Claude, Claude Code, ChatGPT, Cursor, Codex, Gemini, Perplexity Computer and OpenClaw, can connect to https://mcp.arcads.ai and generate on the same credit balance; the credits endpoint reports MCP spend with refunded charges excluded.

Sources

  • Arcads Help Center: How does Arcads pricing work?; Arcads refund policy; How to check the status of your refund or credit reimbursement
  • Arcads external API reference: /v1/credits, /v1/videos

Arcads MCP

Top AI models, one prompt away in Claude, ChatGPT and Grok Bot.

Connect the Arcads MCP once, then generate video and images with top models from your AI assistant.

Add Arcads MCP