CrashSight, developed by Tencent WeTest, is a crash and exception management platform designed for game development teams. The platform captures, reports, and analyzes crashes, ANRs (Application Not Responding), errors, and OOM (Out-of-Memory) issues in real time across mobile, PC, and console platforms.

CrashSight supports iOS, Android, Windows, Linux, PlayStation (PS4/PS5), Xbox, and Nintendo Switch, with native SDK integration for Unreal Engine, Unity, and Cocos. A C interface is available for proprietary engine integration. The SDK is 1.2 MB on Android and 1 MB on iOS with under 2.7 MB runtime memory. A single developer can complete integration in approximately two hours.

Key Features and Capabilities

CrashSight provides a real-time dashboard updated every 10 minutes, displaying crash trends, top issues, and regional statistics. Automated issue classification aggregates exceptions into trackable issues with occurrence counts, affected device metrics, and reporting trends. Each crash report includes detailed diagnostic data: error stacks, trace logs, register information, loaded libraries, and system logs.

The platform offers advanced search across 20+ dimensions including version, device model, OS version, GPU, CPU, region, user ID, and custom fields. Version-to-version quality comparison dashboards and individual user/device timeline tracking help teams identify patterns and prioritize fixes.

Gaming-Specific Intelligence

CrashSight includes capabilities that general-purpose error monitoring tools lack:

— Proprietary OOM detection algorithm using app runtime state machine analysis with device-specific memory thresholds to classify memory crashes that traditional signal-based tools cannot capture

— Emulator and cloud gaming recognition using a big-data approach, achieving 10-15% higher identification accuracy than conventional methods, with separate crash statistics for real devices, emulators, and cloud gaming

— Stack restoration 15x faster through rebuilt symbol table mechanism using .stif format

— Behavioral sequence analysis via Longest Common Subsequence (LCS) to identify recurring user action patterns before crashes

— Country/region-based crash data statistics and custom scene-based crash rate monitoring (e.g., battle, lobby)

Automated feature analysis with customizable rule-based issue tagging reduces repetitive manual work, while predefined platform rules help teams identify common problems faster.

Integrations and Alerting

CrashSight integrates with Jira for one-click bug ticket creation directly from crash issues. Configurable minute-level alerts deliver notifications via Slack, DingTalk, Feishu, WeCom, email, and webhooks. A comprehensive OpenAPI supports custom integrations and automated data retrieval.

Deployment and Compliance

CrashSight is globally deployed with data centers in the United States, Singapore, and Shanghai. All SDK data collection fields are reviewed by the legal department, and overseas data remains in its respective region. CrashSight monitors over 200 million daily active users across its customer base worldwide.

WeTest states that they are is certified under SOC 2 Type 2, ISO/IEC 27001:2013, ISO 9001:2015, ISO/IEC 20000-1, CMMI Maturity Level 3, and TMMI Maturity Level 3.

CrashSight

What is CrashSight?

Categories & Use Cases

Media

Related Products