Recurring Revenue Scaling Rapidly for AI-Media
Key points
- Annual Recurring Revenue (ARR) increased 80% to $30.0M
- ARR gross margin was 84%
- FY26 ARR growth guidance upgraded to >50%
Full summary
AI-Media Technologies Limited (ASX: AIM), a global provider of AI-driven captioning, transcription and translation infrastructure, presented its financial results for the half year ending 31 December 2025 (1H26). The company reported that Annual Recurring Revenue (ARR) increased 80% on the prior corresponding period to $30.0M, with an ARR gross margin of 84%. As a result, the company has upgraded its FY26 ARR growth guidance from 35% to over 50%. Total revenue for 1H26 was $29.8 million, a 6% decrease on the prior corresponding period, reflecting the transition from legacy Services to a recurring revenue model and lower capital sales of end-of-life encoder hardware ahead of next-generation releases. The company's gross margin increased to 70%, driven by the revenue mix shift toward higher-margin recurring revenue. EBITDA was ($1.7M), reflecting continued product investment, including $1.3M in non-cash share-based payments and $2.6M of pre-revenue R&D and product investment. The company had a cash balance of $16.7M at 31 December 2025. AI-Media also announced the launch of its next-generation AI-native encoders under a Hardware-as-a-Service (HaaS) model from April 2026, which is expected to increase recurring revenue intensity and expand lifetime value per customer deployment. The company has withdrawn its previously stated FY29 aspirational revenue and EBITDA targets, and will now focus on disclosing key metrics such as ARR growth, ARR gross margin, and recurring revenue mix.
Guidance
FY26 ARR growth guidance upgraded to >50% (previously 35%).
Outlook
The company's strategic ambition to scale revenue and expand EBITDA over the medium term remains unchanged. The launch of Hardware-as-a-Service is expected to increase recurring revenue intensity and expand lifetime value per customer deployment.
Summary generated by a large language model. Please interpret with care.