Tesla FSD active users surge 56% to 1.48 million in Q2
Tesla Inc. reported strong software growth in Q2, with active Full Self-Driving customers rising 56% to 1.48 million and services revenue hitting $4.58 billion. Meanwhile, Robotaxi paid miles fell 36% to 700,000 as the company expands its service area to gather data for the Cybercab. The Tesla app also saw significant engagement, reaching 10.8 million monthly active users in July.

*this image is generated using AI for illustrative purposes only.
Tesla Inc. is accelerating its transition from a pure vehicle manufacturer to a software-centric enterprise, with second-quarter data revealing a sharp divergence between its autonomous taxi ambitions and its established software ecosystem. The company ended Q2 with 1.48 million active Full Self-Driving (FSD) customers, marking a 56% year-over-year increase. This growth coincided with services and other revenue reaching $4.58 billion, up approximately 50% from the prior year period, driven by record gross profit and gross margin figures.
Conversely, Tesla’s nascent Robotaxi service faced headwinds, recording roughly 700,000 paid miles in the second quarter. This represents a decline of about 36% from the approximately 1.1 million miles logged in the first quarter. Despite the mileage drop, the company expanded its Robotaxi service footprint to additional U.S. metropolitan areas to accumulate driving data for its purpose-built Cybercab, which is currently moving toward production.
Software Ecosystem Expansion
The decline in Robotaxi miles underscores the regulatory and safety hurdles inherent in scaling an autonomous fleet. However, Tesla’s broader software engagement metrics suggest robust adoption among its existing customer base. The Tesla mobile app reached 10.8 million monthly active users in July, according to Similarweb data. This figure reflects a 36.8% increase from a year earlier and a 16.5% rise from June.
Recent app updates have deepened integration with Tesla’s vehicle controls, adding self-driving statistics and expanded functionality. Additionally, Tesla has integrated xAI’s Grok assistant into its vehicles, enabling voice-controlled management of climate and music systems. These enhancements aim to increase the utility and value of vehicles already on the road, reducing reliance on new unit sales for revenue growth.
Revenue Composition Shift
The financial data highlights a strategic pivot where software subscriptions and services are becoming increasingly critical to Tesla’s bottom line. More than 55% of new Tesla deliveries in North America during the quarter included FSD, indicating strong attachment rates for the company’s premium software offering.
| Metric | Value | Change |
|---|---|---|
| Active FSD Customers | 1.48 million | +56% YoY |
| Services & Other Revenue | $4.58 billion | ~+50% YoY |
| Robotaxi Paid Miles | 700,000 | -36% QoQ |
| App Monthly Active Users | 10.8 million | +36.8% YoY |
What the Numbers Show
The divergence between Robotaxi mileage and FSD adoption rates reveals a dual-track strategy. While the high-margin services revenue grew by approximately 50%, the Robotaxi business remains in a data-collection phase rather than a scalable revenue-generating one. The fact that over half of new North American deliveries include FSD suggests that Tesla’s immediate financial upside is tied more to monetizing its existing installed base through software upgrades than to the imminent commercialization of its autonomous taxi fleet. Investors should monitor whether the 56% growth in FSD users can sustain the momentum seen in services revenue as the hardware sales cycle fluctuates.
How might the 36% quarter-over-quarter decline in Robotaxi paid miles impact investor sentiment regarding Tesla's timeline for achieving Level 5 autonomy?
Could the integration of xAI’s Grok assistant create new monetization opportunities beyond vehicle controls, such as in-home energy management or broader AI services?
What regulatory hurdles could emerge as Tesla expands its Robotaxi footprint to additional U.S. metropolitan areas, and how might these affect data collection strategies?

































