# Credits and billing

## Plans

| Plan | Monthly | Yearly | Credits each month |
|---|---|---|---|
| Starter | $29 | $290 | 1,000 |
| Growth | $79 | $790 | 7,000 |
| Scale | $199 | $1990 | 20,000 |

- Paying yearly gets two months free. Every plan has every feature, unlimited products and contacts within fair use, a domain mailbox and 5 web monitors per product. On Starter your public pages show "Powered by Volae"; from Growth up you can turn it off.
- **Enterprise** is custom and adds team members and permissions.
- **Trial:** 14 days, no card, 500 credits.
- **Top-ups:** $10 for 1,000 credits. Top-up credits never expire; monthly credits are used first and don't carry over.
- **Auto top-up** is off by default. When you turn it on you must set a monthly cap; it stops at the cap.
- Prepaid only: Volae never bills overage after the fact.

1 credit = $0.01. AI work is charged at actual usage, so the numbers below for AI items are typical, not fixed; data lookups have fixed prices. Anything over 50 credits gets a quote first.

## Using fewer credits

- **Write with your own Claude Code or Codex.** Context actions and saving what you wrote use no credits. See [Write with your own Claude Code or Codex](own-agent.md).
- **Run tasks on your own model.** AI work on your own endpoint uses no credits; data lookups still do. See [Your own model and data sources](own-model.md).
- **Bring your own data sources.** Calling them uses no data credits.
- **Move up a plan.** If you top up regularly, a higher plan costs less than the lowest plan plus the same top-ups.

## What uses credits

| What | Credits |
|---|---|
| GEO check (per prompt × engine: overseas engines 1, Chinese engines 3, Kimi 7) | 1–7 |
| Voice collection (per platform search: overseas 1, Chinese platforms 4–8, WeChat articles 29, WeChat Channels 54) | 1–8 |
| Chat AI (per resolved conversation) | ~2–3 (at actual usage) |
| Warm email or outreach draft | ~1–2 (at actual usage) |
| Reply drafts & more | ~9 (at actual usage) |
| Agent panel chats and one-off tasks | at actual usage |
| Enrich a contact (work email) | 10 |
| Find a mobile number | 40 |
| Find a personal email | 15 |
| Find people by ICP | ~3 (at actual usage) |
| Keyword research (per run) | 3 |
| Rank check (per keyword × search engine × check) | 1 |
| SEO data: keyword gap and backlink check (at actual usage) | at actual usage |
| Pre-rendered pages for AI crawlers (at actual usage) | at actual usage |
| Content draft | ~7–25 (at actual usage) |
| Cover image | ~9 (at actual usage) |
| Interview transcription (per hour) | 30 |
| Interview notes (per hour) | ~20 (at actual usage) |
| Site audit (per 100 pages) | 10 |
| GEO source tracing (per answer; each cited Douyin or Bilibili video transcribed adds 22) | 2 |
| GEO site review and simulations | at actual usage |
| Social post draft | ~1 (at actual usage) |
| Social simulation | at actual usage |
| Chat replay test (per question, simulated) | ~2–3 (at actual usage) |
| Rival traffic (on demand) | at actual usage |
| Industry search (documents, facts, entities) | 1 |
| Industry records (such as registrations, prices, trials, penalties) | 2 |
| Pre-publish compliance check | at actual usage |
| Industry detection and pack drafting | at actual usage |
| Lookups (per block, at actual cost) | at actual usage |
| Connected account (each, per month) | 1000 a month |
| Extra web monitor (beyond 5 per product, monthly) | 20 a month |

Sorting incoming messages, the coworker's event triage, the daily review and brief, sync, themes, memory and every read action are included in the plan and never use credits. Lookups are priced per block before they run: see [Lookups](lookups.md). Simulation results are estimates: see [Social posts and simulations](social.md).

## When credits run out

Credit-based actions pause; everything included keeps running, and proposals you already have still work. The chat assistant hands new conversations to you instead of answering. You get a notice when you're below 15% and when you're out.

## Payments and read-only

If a renewal payment fails, Stripe retries it automatically. If it hasn't gone through 14 days after the first failure, the workspace turns read-only. Cancelling keeps everything working until the end of the period. Read-only means you can look but nothing runs or changes; top-up credits are kept and your data is kept for 30 days.
