Gemini on Android Auto disappoints despite promising demo

At a glance:

• Gemini on Android Auto has been released but disappoints compared to the impressive demo shown at IO 2025
• Users report issues with voice recognition, conversation continuity, and response times
• While Gemini Live is considered the standout feature, it's still in beta and has its own limitations

The Hype vs. Reality

After seeing what Gemini for Android Auto could do at the IO 2025 demo in Mountain View, many users, including myself, were very excited to get this option in their vehicles. The demo showed an impressive AI-powered voice assistant running on a car head unit that was "almost" on-rails, showcasing what the system could do while still adjusting to off-rails requests without issues. This left many of us excited to enhance our driving experience with what appeared to be a revolutionary new feature.

For those familiar with Gemini Live on mobile, the potential was particularly exciting. On my phone, Gemini Live has become a great way to dive deeper into subjects or get solutions to various problems I face day-to-day. Whether it's troubleshooting software issues or learning about a football player mentioned on the radio, it turns potentially boring moments into engaging conversations rather than requiring endless scrolling through Wikipedia articles or forum posts. It doesn't always offer perfect solutions, but it often provides helpful troubleshooting steps or information that saves time and mental effort.

Current Implementation Issues

Unfortunately, the reality of using Gemini on Android Auto has fallen short of expectations. When my device updated and I noticed Gemini popping up when pressing the voice assistant button on my car steering wheel, my initial happiness quickly turned into frustration and disappointment. It feels like Gemini has been simply hacked onto the Google Assistant rather than being the wholly new experience demonstrated at IO 2025.

I'm not alone in these frustrations. A megathread on Reddit is filled with various complaints from users experiencing similar issues. While many are happy with the new capabilities, a significant number share my concerns about the current implementation. This suggests that while Google has made progress, there's still work to be done to make Gemini on Android Auto truly shine.

Technical Limitations

Several technical issues plague the current implementation. Gemini Live, despite being the standout feature, is still technically in beta and suffers from problems like the annoying voice morphing that has been plaguing Gemini on phones recently. You might start a conversation with one recognizable voice, then launch Live and get someone completely different with all the cadence and response issues that still need fixing.

The connection and response times are also problematic. Depending on your connection, you can often be met with a floating, flashing lilac bar that hangs for a while before anything happens. This breaks the flow of conversation and makes the experience feel sluggish. Additionally, chats don't seem to sync across devices, which means you can't easily pick up a conversation from your car on your phone later—a feature many users would find valuable.

User Experience Frustrations

The user experience has several frustrating aspects. While Gemini Live docks nicely where the music player would ordinarily be on the Android Auto homescreen and looks great, the "ordinary" Gemini appears to have limited response windows while driving or parked. It's not "always listening" as it would be if you initiated Gemini Live directly, which limits its usefulness for quick interactions.

Voice recognition is another significant issue. The regular Gemini isn't great at listening for longer than about 15 seconds at a time, constantly cutting out despite having a solid 5G connection or Wi-Fi. This is particularly problematic when replying to messages while driving, as the system might stop listening mid-dictation, forcing you to re-record or send multiple messages back-to-back.

Design and Functionality Concerns

Gemini introduces unnecessary steps over the previous Android Assistant experience. It doesn't always take into account saved locations on Maps, and it doesn't offer brief retorts to questions or requests as efficiently as the previous assistant did. Many tasks take longer because Gemini is effectively getting in the way and slowing down the process.

The inability to react to display reactions or taps becomes irksome very quickly. For instance, if you ask Gemini to get directions to the nearest McDonald's and there are multiple options, it won't react to your on-screen selection. Instead, it waits for an audio cue or response, slowing things down and breaking immersion. This makes the experience feel janky and unrefined.

The Road Ahead

My biggest disappointment is that after having an in-person demo of Gemini running on Android Auto back at IO 2025, the current product feels like an inferior version. While I'm not frustrated with the handover from the Assistant to Gemini—indeed, it's better at handling imperfect requests like stutters or wrong terminology—the overall experience doesn't match the initial promise.

Gemini for Android Auto was heavily delayed, and the result still feels rushed. I'm sure it'll improve over time, but setting unrealistic expectations at IO is, sadly, a Google trait that many had hoped would be behind us after seeing Gemini running on Android Automotive and Android Auto in person. The current implementation is marginally undercooked and not quite perfect, though with some refinement, it has the potential to be the revolutionary experience that was originally promised.