Skip to main content
Concurrency is the number of active calls your DialNexa account can handle simultaneously. It is one of the most important capacity dimensions for voice AI deployments - especially for batch campaigns, high-traffic inbound lines, and multi-agent setups.

​
What Concurrency Means

If your concurrency limit is 10, your account can have at most 10 calls active at the same time. The 11th call attempt is either queued or rejected, depending on your configuration. Concurrency is counted at the account level across all agents and all workspaces. An account with a limit of 50 that has 3 agents running 20 calls each has consumed 60 out of its 50 concurrent slots - the excess 10 are rejected or queued. Concurrency is not throughput. It is not the number of calls per minute - it is the number of calls happening at the same moment. A 3-minute average call duration with a concurrency of 10 allows approximately 200 calls per hour to complete (10 concurrent × 60 minutes / 3 minutes average). For higher throughput, you need higher concurrency.

​
Default Limits by Plan

PlanConcurrent callsSuitable for
Starter5Testing, pilot deployments, low-traffic inbound
Professional25Growing businesses, moderate campaign volumes
Business100High-traffic deployments, active outbound campaigns
EnterpriseCustomLarge-scale deployments, thousands of concurrent calls
These are default limits. You can request a temporary or permanent limit increase for any plan - see below.

​
What Happens When the Limit Is Reached

When a new call attempt (inbound or outbound) would exceed the concurrency limit: For outbound calls (batch campaigns): The call is queued. The platform continues trying to dial the contact once a concurrent slot frees up. Queue depth is limited - if the queue fills, excess calls are marked as skipped in the campaign report and can be retried in a subsequent campaign run. For API-initiated outbound calls: The API returns a 429 Too Many Requests error with a retry-after header indicating how many seconds to wait before retrying. For inbound calls: The call is rejected at the carrier level. The caller hears a busy signal or a “number unavailable” message, depending on the carrier. For inbound call centers, hitting the concurrency limit means callers cannot reach you - this is a service outage equivalent.
Hitting your concurrency limit on an inbound line is a customer-facing failure. Monitor your concurrent call count and request a limit increase well before you expect to approach your cap.

​
Monitoring Concurrent Call Usage

Real-time: The dashboard’s Monitoring → Live View shows the current number of active calls across all agents in real time. Historical: The Monitoring → Analytics page shows a chart of concurrent calls over time. Use this to understand your peak concurrency and whether you are approaching your limit during high-traffic periods. Alerts: Configure a concurrency alert in Settings → Billing → Alerts to be notified when your concurrent call count exceeds a threshold (e.g., alert me when concurrent calls > 40 on a Business plan with a 100-call limit).

​
Requesting a Limit Increase

​
Temporary Increase (for campaigns)

If you are planning a time-limited batch campaign that requires higher concurrency than your plan allows:
  1. Contact [email protected] at least 48 hours before the campaign launch
  2. Specify: your account, the requested concurrency level, and the campaign window (start date/time, estimated duration)
  3. Support will temporarily raise your limit for the campaign period at no additional charge (within reasonable bounds)

​
Permanent Increase

To permanently increase your concurrency limit:
  1. Go to Settings → Plan and click Upgrade Plan, or
  2. Contact [email protected] for Enterprise plans with custom limits
Permanent increases may involve moving to a higher plan tier or negotiating a custom limit for Enterprise accounts.

​
High-Concurrency Use Cases

Some use cases require significantly higher concurrency than standard plans provide:
A business sending appointment reminders to 100,000 customers within a 4-hour window needs to dial roughly 7 contacts per second. To achieve 50% connect rate with 3-minute average calls, you need approximately 500-600 concurrent call slots during the campaign peak.
Political and civic outreach campaigns often target millions of voters in compressed timeframes. These campaigns can require 1,000+ concurrent calls.
Banks and NBFCs sending fraud alerts or loan approval notifications need to reach customers quickly and simultaneously, requiring very high short-burst concurrency.
Insurance companies running renewal campaigns at policy expiry seasons need sustained high concurrency over weeks.
ISVs serving many clients simultaneously where each client’s agents may be running campaigns at the same time.
For all these scenarios, contact [email protected] to discuss custom concurrency limits and Enterprise pricing. We provision custom infrastructure for very high-concurrency deployments.

​
Concurrency and Cost

Increasing concurrency does not directly increase your per-minute call cost - you pay the same per-minute rate regardless of how many concurrent calls are active. Concurrency limit increases on higher plan tiers come with the higher plan’s per-minute rate, which is typically lower. For Enterprise customers, very high concurrency tiers are priced based on committed call volume, typically resulting in the lowest per-minute rate.

​
Concurrency for Web Calls

Web calls (browser-based calls via the React SDK) count against the same concurrency limit as phone calls. If your account has a limit of 25, and 20 web callers are simultaneously connected to your agent, only 5 concurrent phone call slots remain. Plan your concurrency limit to account for all call channels you use.