Solid Performance Paving the Way for High-Quality Growth—Maintaining Buy Rating

Overview

UBS released an earnings commentary on iFlytek’s 2025 and first-quarter 2026 results. The report noted that the company’s performance during the period was solid and is moving toward a higher-quality growth phase: 2B and 2C businesses grew significantly faster than 2G businesses, and cash flow improved markedly. Technically, the company has strengthened its core advantage in training large models using domestically produced computing power, accelerating commercial monetization of these models. UBS adjusted its earnings forecast and, based on the DCF model, maintained a target price of RMB 60 and a 'Buy' rating.

Core views

Performance and Growth Structure: In 2025, iFlytek’s revenue reached RMB 27.105 billion and net profit RMB 839 million, up 16% and 49% year-on-year respectively, with Q4 2025 revenue growing by 19%. In the first quarter of 2026, revenue was RMB 5.27 billion (up 13% year-on-year), and net loss was RMB 170 million (narrowed by 12% year-on-year). In terms of growth quality, 2B and 2C segments grew significantly faster than 2G, and the proportion of 2G revenue dropped from 29% in the previous year to about 26% in 2025. Meanwhile, the company’s cash collections in 2025 and 1Q26 reached RMB 27.4 billion and RMB 5.7 billion respectively, both exceeding revenue for the same periods, reflecting continuous improvements in operational quality through better project selection and credit management. Large Model Technology and Commercialization Progress: The company reaffirmed its core advantage in training large models on fully domestically produced computing infrastructure and plans to release a new-generation flagship large model equipped with more advanced domestic computing power (such as Ascend 950) in October. On the commercial side, the number of large model developers on the open platform grew by 124% year-on-year to 2.29 million in 2025, and revenue from APIs and MaaS services reached RMB 385 million, a 263% increase year-on-year. Education Business and Overseas Expansion: In educational hardware, the company continues to gain market share in the mid-to-high-end learning device market above RMB 5,000, launching the new T90 model in 1Q26. Although sales growth slowed in 1Q26 due to the suspension of government subsidies, management remains confident in its penetration potential (currently only 7-8%). Overseas expansion is going smoothly; in 2025 (excluding mainland China), overseas revenue exceeded RMB 1 billion, and 1Q26 still maintained triple-digit growth, driven mainly by the multilingual capabilities of the Spark large model boosting AI hardware sales.

Analysis framework

UBS conducted its analysis along the main line of 'performance verification—structural analysis—technological barriers—commercial monetization—valuation.' First, it confirmed the fundamentals’ stability by comparing actual results with forecasts. Then, it dissected the business structure, using comparisons of 2B/2C and 2G growth rates and collection data to demonstrate the company’s transition toward high-quality growth. On the technology front, it emphasized the core barrier of ‘training on domestically produced computing power’ and tracked indicators such as the number of developers and API revenue to validate the commercial implementation of large models. Finally, it used a three-stage DCF model for valuation, adjusting revenue forecasts for each segment to update the financial model and derive the target price.

Methodology notes

• Valuation MethodDCF Cash Flow Discounting

  Three-stage DCF (Discounted Cash Flow) Valuation Method

  DCF calculates enterprise value by forecasting future free cash flows and discounting them to the present. UBS adopted a three-stage model, corresponding to a high-growth period, a medium-term growth period, and a perpetual growth period. This method is suitable for generative AI companies in early high-growth stages with low short-term cost-revenue matching. Since there are no directly comparable listed peer models, DCF becomes a more reasonable valuation choice.

• Industry/Industrial Analysis FrameworkSupply-demand framework

  Application of the Supply-Demand Framework in 2G Project Screening and Collection

  The report observed the company’s proactive screening of 2G projects (focusing more on project quality) and its cash collection performance, indirectly reflecting optimization of industry supply-demand matching. This illustrates the analytical logic that when external demand (such as government budgets) is uncertain, the company improves overall operating asset quality by contracting and optimizing the supply side.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• iFlytek (002230.SZ)

  This report directly covers the target stock and benefits from the commercialization of large models and the transition toward high-quality growth.

  Strengths

  Core advantage in training large models on domestically produced computing infrastructure; increasing market share in mid-to-high-end educational hardware; overseas business maintaining triple-digit growth; improved cash flow management capability

  Risks

  Tightening government budgets, economic slowdown dragging down corporate IT spending, intensified competition leading to lower-than-expected sales, ongoing investment in large models putting pressure on profits and slow monetization, heavy GPU capital investment impacting cash reserves

Key data

• 2025 RevenueRMB 27.105 billionUp 16% year-on-year
• 2025 Net ProfitRMB 839 millionUp 49% year-on-year
• 1Q26 RevenueRMB 5.27 billionUp 13% year-on-year
• 1Q26 Net LossRMB 170 millionNarrowed by 12% year-on-year
• 2B and 2C Business Growth2025 +20%, 1Q26 +26%Faster than 2G business growth
• Large Model API and MaaS RevenueRMB 385 millionUp 263% year-on-year in 2025
• Number of Developers2.29 millionUp 124% year-on-year in 2025
• Target Price and Implied ReturnRMB 60.00 / 24.8%Maintaining target price, implying an upside potential of 24.8%

Impact & implications

UBS believes that iFlytek is in a transitional phase from pursuing scale to pursuing quality. The high growth of 2B and 2C businesses and the technological barrier of training large models on domestically produced computing power have laid the foundation for the company’s medium- and long-term development. Although it faces short-term pressure from 2G budget constraints and the impact of R&D spending on profitability, the company’s operational resilience and cash flow are improving through optimized project quality and enhanced commercial monetization capabilities. Maintaining the target price implies that the institution sees room for further upside in the current valuation.

Risks

• Further tightening of government budgets
• Continued economic slowdown increasing pressure on corporate IT spending
• Intense competition causing slower-than-expected sales growth
• Ongoing investment in large models putting pressure on profits and slow monetization, resulting in poor returns
• Heavy GPU capital investment potentially putting pressure on the company’s cash reserves

What to watch

• Performance of the new-generation flagship large model to be released by the company in October
• Progress in deploying more advanced domestic computing power such as Ascend 950
• Improvement in the penetration rate of educational hardware products
• Further changes in the revenue mix between 2G and 2B/2C businesses