TRTC vs Vonage Video API: Web Video Quality Benchmark Under Weak Networks

Video quality under degraded network conditions is one of the most important factors when evaluating a real-time communication SDK.

Under ideal network conditions, many video SDKs can deliver similar results. The differences become much clearer when users encounter packet loss, bandwidth constraints, or high jitter — conditions commonly experienced on congested Wi-Fi, VPN connections, and bandwidth-constrained networks.

In this benchmark, we compared Tencent RTC (TRTC) and Vonage Video API under controlled weak-network conditions.

Both SDKs used identical video sources, matching configurations, and the same network impairment conditions.

Key Results at a Glance

Test configuration:

The primary benchmark results were:

Across the six weak-network scenarios, TRTC delivered an average effective pixel rate of 7.13 Mpx/sec, compared with 2.57 Mpx/sec for Vonage.

In this benchmark, that represents approximately 177% higher average effective pixel rate.

TRTC also remained at the configured 640×480 resolution for 100% of the weak-network test period, while Vonage remained at that resolution for an average of 33% of the time.

Average measured recovery time was 1.0 second for TRTC and 6.8 seconds for Vonage.

Test Methodology

Both SDKs were tested under identical controlled conditions.

Frame-rate measurements refer to frames actually rendered at the receiving browser, rather than frame-rate values reported internally by the SDK.

Web Test Setup

Test Scenarios

Each scenario followed the same overall sequence:

Establish baseline → apply weak-network conditions → remove restrictions → observe recovery

These conditions represent unstable or constrained network environments, including congested connections and bandwidth-limited networks.

Test constraints

This report presents TRTC performance with the “Ensuring video smoothness” preference, i.e. the setting that prioritizes video smoothness over image clarity. This is an account-level setting that is disabled by default; to reproduce the test, contact TRTC Technical Support​ to have this setting provisioned for your account.

Web Video Quality Results

Effective Pixel Rate Under Weak-Network Conditions

For the Web benchmark, we use effective pixel rate to evaluate video quality.

Effective pixel rate combines two important dimensions of the received video:

Effective Pixel Rate = Frame Rate × Resolution

The result is expressed in megapixels per second (Mpx/sec).

This metric is useful because an SDK can respond to weak-network conditions in different ways. For example, it may: lower the frame rate while retaining the original resolution; lower the resolution while maintaining a higher frame rate; reduce both frame rate and resolution.

Looking only at FPS can therefore hide changes in picture detail.

Effective Pixel Rate by Scenario

Effective Pixel Rate Difference

The largest differences appeared in the uplink packet-loss and jitter tests.

The 500 Kbps bandwidth-cap test was the closest result, with TRTC at 4.21 Mpx/sec and Vonage at 4.28 Mpx/sec.

Why Use Effective Pixel Rate Instead of FPS Alone?

Under bandwidth constraints, real-time video SDKs need to balance picture detail and motion smoothness.

The test results suggest different behavior between the two SDKs.

Because of this difference, comparing frame rate alone would not completely represent the received video quality.

Effective pixel rate combines both resolution and frame rate into one metric, making it easier to compare the total amount of visual information delivered per second.

Resolution Retention Under Weak Networks

Another important metric is how often the received video remains at the configured 640×480 resolution when network conditions deteriorate.

Across the Web weak-network tests, TRTC remained at 640×480 throughout the measured weak-network periods.

Vonage remained at 640×480 for an average of 33% of the time, with the remaining period spent at lower resolutions such as 480×360 or 320×240.

Why Resolution Retention Matters

Resolution becomes particularly important in applications where users need to inspect visual details.

At lower resolutions such as 320×240, fine visual information becomes significantly harder to distinguish.

Recovery Time After the Network Returns to Normal

Weak-network performance is only one part of the user experience.

Another important question is:

How quickly does video quality return after network restrictions are removed?

In this benchmark, recovery was defined as reaching 90% of the baseline level and maintaining that level for three consecutive seconds.

A recovery time greater than 30 seconds was treated as not recovered within the observation window.

Recovery Time by Scenario

The recovery results vary by scenario.

Vonage recovered faster in several individual bandwidth or downlink scenarios.

The largest differences appeared under 600ms jitter:

Because recovery behavior varied considerably across network conditions, average recovery time should be considered together with the scenario-level results.

Frame Rate Over Time Under Weak-Network Conditions

The tables above summarize the benchmark numerically.

However, averages do not fully show when the frame rate drops, how much it fluctuates during the impaired period, or how quickly it stabilizes after the restriction is removed.

The original test includes time-series frame-rate curves for the following scenarios:

• Y-axis: receiver frame rate (fps)
• X-axis: time within scenario
• Orange-shaded region: uplink restrictions active
• Red-shaded region: downlink restrictions active
• Green-shaded region: recovery observation window

Figure: TRTC and Vonage Video API receiver frame-rate changes during baseline, active network impairment, and recovery periods.

What These Results Mean in Practice

The importance of each metric depends on the application.

Some products depend heavily on picture detail, while others place more weight on smooth motion or recovery from unstable networks.

Interactive Live Streaming

1:1 Video Calls

Conversational AI and Video Agents

Stable video input may also be relevant for AI-powered voice and video applications.

Weak-network conditions can introduce:

These effects may be relevant for applications using incoming video for: facial analysis; visual context processing; gesture recognition; emotion-related analysis; multimodal AI interaction.

The benchmark data can therefore also be considered when evaluating RTC infrastructure for AI-driven video applications.

Observed SDK Behavior Under Bandwidth Constraints

The benchmark results suggest different adaptation patterns between TRTC and Vonage.

These observations are based on measured output and do not represent confirmed information about the internal algorithms or implementation of either SDK.

TRTC Behavior Observed in This Test

Under constrained bandwidth, TRTC remained at the configured 640×480 resolution while frame rate decreased.

For example, under the most constrained Web scenarios, effective pixel rate dropped as frame rate adapted to available network capacity.

At a 640×480 resolution, these results correspond to reduced frame rates compared with the baseline while retaining the configured image resolution.

Vonage Behavior Observed in This Test

Vonage spent a larger portion of the weak-network period at reduced resolutions such as 480×360 or 320×240.

Across all weak-network scenarios:

This indicates that the received resolution changed more frequently under the tested network constraints.

Which adaptation strategy is preferable depends on the application.

Complete Web Benchmark Data

For easier comparison and machine-readable reference, the main Web benchmark results are consolidated below.

Resolution Summary

Benchmark Summary

Under the Web test configuration used in September 2026, the two SDKs showed different behavior across packet loss, bandwidth constraints, and jitter.

The main measured findings were:

TRTC averaged 7.13 Mpx/sec effective pixel rate, compared with 2.57 Mpx/sec for Vonage.

This corresponds to approximately 177% higher average effective pixel rate for TRTC in these six weak-network scenarios.

TRTC remained at 640×480 for 100% of the measured weak-network period, compared with 33% for Vonage.

Average measured recovery time was 1.0 second for TRTC and 6.8 seconds for Vonage.

The effective pixel rate results were closest under the 500 Kbps bandwidth-cap scenario, at 4.21 vs 4.28 Mpx/sec.

The largest recovery-time differences appeared under 600ms jitter.

Recovery performance varied by scenario, with Vonage recovering faster in some packet-loss and bandwidth-cap cases.

These results apply specifically to the devices, browser version, SDK configuration, video configuration, and network conditions described in this benchmark.

Performance may vary under different browser versions, hardware configurations, network routes, SDK releases, and application settings.

Test Information

All benchmark data reflects the test configuration described in this article.

Results may vary under different network conditions, device combinations, browser environments, and SDK configurations.

For questions about this benchmark or Tencent RTC, visit Tencent RTC Contact.