Can Apple block the number? Claw Messenger limits, ban risk, and reliability
Apple does not offer a public iMessage API for servers, so every hosted iMessage API, Claw Messenger included, works without Apple's involvement, and Apple can restrict a sending number that behaves like spam. Claw Messenger keeps its lines healthy by limiting how many new people an account can start conversations with, capping contacts per plan, and refusing bulk use.
The short answer
We are not a blast tool, and that is a feature: the limits that stop bulk sending are what keep conversations delivering. If you need high-volume outbound to new contacts, Claw Messenger is the wrong fit.
- Best for
- Agents that hold ongoing conversations with people who expect to hear from them.
- Not for
- Bulk marketing, cold outreach, or campaigns to purchased lists.
At a glance
| Monthly messages | Base 250 · Growth 2,000 · Plus 6,000 · Pro 15,000, inbound and outbound |
|---|---|
| New-recipient safeguard | 20 distinct outbound recipients per rolling 24 hours by default (self-serve). At the limit, only first-time recipients pause. |
| Registered contacts | Base 1 · Growth 5 · Plus 20 · Pro 50 |
| Agency line limits | 6,000 messages per line per UTC day; 50 new conversations per line per rolling 24 hours |
| Warm-up period | None to wait out; daily new-recipient limits apply from day one |
| WebSocket connections | Up to 20 at once per account |
| Requests per second | No published guarantee |
| Delivery time | No published figure |
| Status | Live status and 90-day history at /status |
Is iMessage automation against Apple's terms? Can the number be banned?
Apple does not offer a public iMessage API for server applications. Apple Messages for Business is a separate, Apple-run customer-support channel. Every hosted iMessage API, and every Mac bridge such as BlueBubbles, works without Apple's involvement. Apple can restrict a sending number that sends like spam, and no provider can promise that it never will.
What that means for you on Claw Messenger:
- Your own Apple account is not involved. We manage the iMessage line upstream, so there's no Mac, no SIMs, no iCloud, and no BlueBubbles on your end. You never sign in with your Apple ID.
- We run line health. The limits below exist to keep sending numbers looking like what they are: people in conversation with an agent.
- Messages keep a fallback. When a recipient can't get iMessage, the message goes as RCS or SMS instead.
- Your account survives a number change. If a sending number has a problem, your API key, contacts, and history stay as they are. See what happens if a line has a problem.
If you need Apple's own approval and a verified brand in the thread, Apple Messages for Business is the better fit. Read our Apple Messages for Business guide.
What is the new-recipient safeguard?
On self-serve plans, every distinct phone number your account sends to counts once in a rolling 24-hour window, including existing contacts. Repeated sends to the same person count once. Contacts who only text you don't count.
| Distinct outbound recipients in 24 hours | What happens |
|---|---|
| 0 to 9 | Normal use. |
| 10 to 19 | We are notified internally. Sending continues and you see no change. |
| 20 (default) | Sends to people your account has never messaged return 429 Too Many Requests until the rolling count falls. Existing conversations continue. The account is also flagged for a human review. |
Accounts that review finds sending spam can be suspended under the terms. The same rules are on plans and limits.
Do I need to warm up the number?
No. There is no warm-up period to wait out before your agent can start conversations. The daily limits on new recipients apply from the first day, and they do the pacing a warm-up would. Good habits still help:
- Let people text your agent first when you can.
- Introduce new contacts a few at a time, not all on one day.
- Keep messages conversational and expected. Don't paste the same text to a list.
- On a
429, don't retry in a tight loop. HonorRetry-After, back off with jitter, and queue the message.
What happens when you hit a limit?
| Limit | What happens | When it clears |
|---|---|---|
| Monthly message allowance | Sends fail with Monthly message limit reached | Next monthly period, or upgrade |
| New-recipient safeguard (self-serve) | Sends to first-time recipients return 429; existing conversations continue | As the rolling 24-hour count falls |
| Registered contacts | Registering a new contact is refused with phone_route_limit_reached | Release a contact, or upgrade |
| Agency messages per line per day | Outbound sends on that line fail with Agency daily message limit reached | Midnight UTC |
| Agency new conversations per line | Sends to never-seen numbers return agency_line_new_conversation_limit_reached (retryable) | As the rolling 24-hour window moves |
How do Agency line limits work?
Agency limits are counted per sending line, across all traffic on that line, not per account. Each line handles up to 6,000 messages per UTC day and may start up to 50 new conversations in a rolling 24 hours. A conversation is new when the line has never exchanged a message with that number in either direction, so contacts who text first don't count. Replies in existing conversations are never blocked by these limits. Spread client onboarding across days or across lines. Details are in the Agency docs.
What happens if a sending line has a problem?
- Self-serve: if the sending number can't send, messages move to a backup number automatically so conversations keep going. Recipients then see that number. If a number has to be replaced, conversations move to the new number and recipients may get a short note that the agent is writing from a new number.
- Agency: while a line is not healthy, it can't start new conversations, and existing conversations continue. A send the line can't deliver returns an error instead of quietly going out from a different number.
- Replies wait for your agent. Inbound replies are stored before delivery, so an agent that was offline receives them when it reconnects.
- You can see it. The status page shows live API and message delivery status, 90 days of daily history, and past incidents. If a message looks stuck, email clawmessenger@mail.tin.computer with the delivery status and approximate send time.
How do other iMessage APIs limit new contacts?
| Provider | Starting conversations | Entry price | Source |
|---|---|---|---|
| Claw Messenger | Self-serve: 20 distinct outbound recipients per rolling 24 hours, then first-time recipients pause. Agency: 50 new conversations per line per 24 hours. | $5/mo (Base) | Plans and limits |
| Sendblue | Free Sandbox: no outbound. AI Agent: inbound-first, 1,000 inbound contacts/day. Full outbound on Enterprise. | $0 sandbox; $100/mo per line | sendblue.com/pricing |
| Blooio | Starter 5 new contacts/day; Commercial Shared 15/day; own-number commercial plan unlimited; Inbound plan reply-only. | $39/mo | blooio.com/pricing |
| LoopMessage | Shared sender can't start conversations. Starting them is a paid add-on on a per-sender plan, after a warm-up period and with recent contact consent. | $20/mo shared sender; $59.99/mo per sender | loopmessage.com/pricing |
| Photon | Business and Enterprise: up to 50 new contacts/day. Free and Pro: not published. | Free (up to 10 users) | photon.codes/pricing |
If your agent mostly answers people who text first, an inbound-first plan elsewhere may suit you. If you need many new outbound conversations a day from one number of your own, a per-line plan from Blooio or Sendblue Enterprise fits better than Claw Messenger.
Why Claw Messenger is not a blast tool
The terms rule out spam, unsolicited messages, and bulk marketing. That is a product choice, not fine print. Lines that carry real conversations keep delivering, and that is what your agent depends on. If your plan is a campaign to a list, use a marketing SMS provider instead.
Common questions
Is using an iMessage API against Apple's terms? Can the number be banned?
Apple does not offer a public iMessage API for servers, and no hosted iMessage API, Claw Messenger included, is an Apple product. Apple can restrict a sending number that behaves like spam. Claw Messenger limits new recipients and contacts and refuses bulk use to keep its lines healthy. No Apple ID of yours is involved, so your personal Apple account is not the one at risk.
Do I need to warm up a number before my agent can start conversations?
There is no warm-up period to wait out. The daily limits on new recipients apply from the first day, and they pace new conversations for you.
What happens if my agent texts too many new people in a day?
On self-serve plans, sends to people your account has never messaged pause once 20 distinct outbound recipients are reached in a rolling 24 hours. They return 429 Too Many Requests until the count falls. Existing conversations continue.
Is there a requests-per-second limit?
Claw Messenger does not publish a requests-per-second guarantee. Pace sends conservatively, honor Retry-After on 429 responses, and treat API responses as the source of truth.
How fast are messages delivered?
Claw Messenger does not publish a delivery-time figure today. Each send returns a status, and delivered, read, and failed events follow when they are reported.
Where can I check for outages?
The status page at clawmessenger.com/status shows live API and message delivery status with 90 days of history and past incidents.
Can I send marketing campaigns with Claw Messenger?
No. Spam, unsolicited messages, and bulk marketing are against the terms. Claw Messenger is for conversations people expect.
Related pages
Building an agent that talks with people who expect to hear from it? Start with one contact.
Start the 7-day trial