How throughput works
About a 3-minute read.
"Throughput" is how fast you can send, how many message segments per second and per day. If your agent fires a burst and the messages drip out instead of going at once, you're hitting a throughput ceiling. Here's what actually sets that ceiling on AgentMessage, because it's not what older 10DLC guides describe.
Two ceilings, and you get the lower one
Your effective send rate is the lower of two limits:
1. AgentMessage's send-rate tier (rises automatically with usage)
AgentMessage applies a per-account send-rate tier, a ceiling on segments per second and per minute. This tier advances automatically as your usage grows. You don't apply for it, buy it, or configure it. A brand-new account starts on the entry tier; as you send more, you move up to higher per-second and per-minute ceilings.
For most agent builders, especially early on, this is the ceiling you'll actually feel, and it lifts on its own as your program scales.
2. The carrier caps on your brand
On top of that, each carrier enforces its own limits based on your brand's standing (its verification/trust, use case, and that carrier's rules). These caps exist regardless of how fast AgentMessage would otherwise let you go.
Your real ceiling at any moment is whichever of these two is lower. Sending faster than the cap doesn't push more through, it just queues, and the queue drains at the allowed rate.
Adding more phone numbers does not make you faster
This catches people out. Throughput is enforced at the brand/campaign level, not per-number. Spreading the same traffic across more numbers splits the same throughput across more lines, it doesn't raise the total. More numbers are useful for other reasons (per-region presence, separating workloads), not for raw speed. (And rotating numbers to dodge limits, "snowshoeing", is exactly what carriers watch for and suspend over.)
If you're hitting a ceiling
- Confirm what's actually capped. Is the whole account queuing on every send (AgentMessage send-rate tier), or are sends to one carrier specifically failing late in a large batch (a carrier daily cap)?
- Give it room to grow. The AgentMessage tier rises with usage; a temporary burst ceiling on a young account often resolves as you scale normally.
- Separate time-critical traffic. Putting 2FA or fraud-style alerts on their own campaign keeps them from waiting behind a big marketing batch.
- For genuinely high-volume needs, talk to us. Lifting carrier-side caps further can involve external brand vetting (which produces a carrier trust score) or carrier-level capacity reviews for high-volume senders. These aren't self-serve buttons in the dashboard, contact us with your volume needs and we'll coordinate the right path with the carriers.
What won't help
- Switching messaging providers. Carrier caps follow your brand at the registry level, not your vendor.
- Splitting one brand into several. Expensive, risky, and each new brand has to earn its own standing from scratch.
- Just sending harder. The network throttles; flooding only grows the queue.
A note on what you'll see
Brand and campaign detail pages reflect your standing and status. If you ever need to understand a specific delivery slowdown, contact support and we can look at where the throttling is coming from.