Agora, Inc. Q3 FY2025 earnings call

• Q3 saw fourth consecutive quarter of GAAP profitability with $2.7 million net profit and 7.8% GAAP net margin. • Significant investment in conversational AI, including launch of conversational AI engine 2.0 with AI-powered noise suppression, etc. • Launched conversational AI studio, benchmark and orchestration platform, and open-source time framework. • Expanded global distributed real-time assurance cloud across key regions. • Q3 R&D expenses $13.8M, down 52.8% Y/Y; sales and marketing expenses $6.5M, down 5.6% Y/Y; G&A expenses $5M, down 48.4% Y/Y.

Total revenue in Q3 reached $35.4 million, up 12% year-over-year. Core revenues were $18.2 million in Q3, up 15.9% year-over-year and flat quarter over quarter. Shunghwa revenues reached RMB 122,400,000 in Q3, up 8.4% year over year and 6% sequentially. Dollar-based network retention rate is 108% for Agora and 90% for Shunghu. Gross margin for the third quarter was 66%, slightly decreased 0.7% year over year and 0.8% sequentially.

• Expect total revenues for 2025 to be between $37 million and $38 million. • Q4 expected to achieve GAAP operating profit breakeven. • Target GAAP operating profit for full year 2026; GAAP net profit expected to show big improvement over 2025, with some uncertainty due to potential interest rate cut but expecting year-over-year net income improvement over 2025.

Q: Regarding demand outlook, key trends in domestic and international markets and downstream sectors driving growth; update on AI application revenue contribution; profitability outlook for Q4 and FY '26.

A: Tony Zhao discussed China demand recovery in social, education, IoT; US/international live commerce growth; clarified conversational AI vs voice AI, with call centers, education, companionship toys as advanced use cases, expecting conversational AI revenue in first half of next year. Jingbo Wang said Q4 expected GAAP operating profit breakeven, 2026 target GAAP operating profit, net profit improvement over 2025.

Q: Q3 revenue above guidance midpoint drivers; strongest AI downstream applications, adoption trend for AI companionship toys.

A: Jingbo Wang mentioned strong demand from US international live commerce, China summer vacation and IoT; Tony Zhao talked about call center pipelines, RoboPong's success with AI companionship toys and other toy manufacturers integrating solutions.

Q: Sequential growth trend for AI-related usage, timeline for meaningful scale; potential strategy extension into edge-side inference or security.

A: Tony Zhao said conversational AI usage increased over 150% Q/Q; infrastructure for real-time inference services as an expansion opportunity.