Apple's latest iOS 27 beta 5 update continues to refine Siri, and the newest change expands the voice customization tools that were first introduced in earlier betas. Pace and Expressivity controls are no longer limited to two new American-accent Siri voices. Instead, all American voices and all existing British accent voices now support these settings on compatible devices.
The update is part of a broader effort to make Siri feel more natural and personal. Rather than relying on a one-size-fits-all voice, Apple is increasingly giving users the tools to shape how Siri sounds. The fact that the controls are arriving at the beta stage suggests that Apple is listening to user feedback and is willing to adjust course before iOS 27 is released to the public.
What's New in iOS 27 Beta 5
When iOS 27 first introduced the new Siri voice options, the ability to adjust pace and expressivity was restricted to two newly created voices with an American accent. This created a somewhat confusing situation: users who preferred British English voices, or who wanted one of the older American voices, had no way to customize the delivery style.
With beta 5, Apple has removed that limitation. Users can now apply the pace and expressivity sliders to any American voice, as well as all the existing British accent voices. This is a significant change because it broadens the feature from an experimental addition to a general voice customization option.
Apple states that there are now eleven supported voices in total that allow pace and expressivity adjustment. On a modern iPhone, these eleven voices are divided into eight American options and four British options. Some of these additions have been in Siri for years and are familiar to long-time users, but they now behave differently because the neural synthesis engine can adapt them in real time.
Understanding Pace and Expressivity
Pace controls how quickly or slowly Siri speaks. It is a straightforward setting that can make a big difference in usability. Some people prefer a faster voice because they are trying to get information quickly, while others prefer a slower, calmer pace, especially when listening in a noisy environment or when they are not native English speakers.
Expressivity is more subtle. It affects the emotional tone and emphasis of Siri's speech. A more expressive voice might sound more excited, sympathetic, or forceful depending on the context. This is especially useful for responses that include greetings, warnings, or interactive phrasing. With five levels for each slider, users have a decent range of choices. The sliders are located in Settings, Siri, and Voice, directly under the selected voice.
Hardware Requirements and Supported Devices
Not every iPhone will be able to take advantage of these features. Apple has said that the new Siri voice synthesis relies on a powerful on-device model. The implementation requires an iPhone 17 Pro, iPhone 17 Air, iPad Pro with M4 or newer, and Macs with an M3 chip or newer and at least 12 GB of RAM.
These are notable requirements. They mean that some devices that support iOS 27 generally will not support the new voice model. For example, an iPhone 15 Pro might be able to install iOS 27, but it will not have access to the same voice output quality or the pace and expressivity controls if the feature is tied to the newer on-device model.
It also explains why Apple has been careful to frame the feature as an on-device AI advancement rather than a simple user interface update. The model is large enough that it needs the memory bandwidth and compute power of Apple's latest silicon. Users with older hardware will continue to get the previous Siri voice experience, which is still serviceable but not as flexible.
Siri AI and the Voice Synthesis Distinction
It is important to separate Siri AI from the new voice synthesis. Siri AI, Apple's larger generative assistant update, will require an iPhone 15 Pro or newer. That means a range of devices can run Siri AI even if they cannot run the newest voice model. Those devices will get Siri AI features but will use the older voice synthesis.
Apple has said that Siri AI will launch in beta later this year. When that happens, users should expect more capable language understanding, better context handling, and deeper integration with apps and personal information. However, the way Siri voices are generated will remain tied to the hardware requirements of the voice model.
This split may seem unusual at first, but it makes sense from a technical perspective. Siri AI is largely about understanding and generating text. Voice synthesis is a separate challenge that involves converting that text into audio. The two systems involve different neural networks and memory footprints.
The Shift to Neural Voice Synthesis
Siri's original voices were built using unit selection synthesis, which captured a large library of recorded speech segments. The system would stitch together small pieces of audio to form complete sentences. This worked reasonably well for short commands, but it could sound robotic, especially with long sentences or uncommon words.
Apple has gradually moved toward neural synthesis, in which a deep learning model generates audio directly from text. Neural models can capture rhythm, stress, and intonation more accurately. They also allow for continuous adjustment of speed and emotional expression, which is why the pace and expressivity sliders have become possible.
Because these models are full of learned parameters, they can be computationally expensive. Apple's solution has been to make them efficient enough to run locally on modern Apple silicon. This is also why the new voices are only available on newer devices. The company is not artificially limiting the feature; the technology genuinely needs the resources.
Privacy and On-Device Processing
One of the key benefits of the new voice synthesis pipeline is privacy. When voice generation happens on-device, Siri does not have to send the text of the response to Apple servers just to produce the audio. This reduces the amount of personal data leaving the phone and helps Apple maintain its privacy promises.
On-device synthesis also improves reliability. It works in airplane mode, in areas with poor connectivity, and in situations where users do not want their device to make a network request before Siri can answer. The local model is fast enough to generate speech without any noticeable delay, which makes Siri feel more responsive.
This approach is consistent with Apple's broader direction across iOS 27, which has emphasized on-device intelligence. Many of the new AI features are designed to work without cloud assistance, and voice is no exception.
How to Use the New Controls
To customize Siri's voice on a compatible iPhone or iPad, open the Settings app and choose Siri. From there, select Voice and choose a voice that supports pace and expressivity. A pair of sliders will appear below the voice preview. The first slider controls pace, and the second controls expressivity.
Each slider has five supported levels. Moving the pace slider to the right makes Siri talk faster, while moving it to the left slows things down. The expressivity slider follows a similar pattern, with the right making Siri sound more emotionally involved and the left producing a flatter, more neutral delivery.
Users can test different combinations by triggering a Siri response or using the available preview. Because each voice remembers its own settings, it is easy to create different assistants for different purposes. For instance, a user might choose a faster, less expressive voice for quick reminders, and a slower, more expressive voice for driving directions or casual conversation.
Accessibility and Assistive Benefits
The expanded voice customization is also relevant for accessibility. Users with certain visual impairments or reading difficulties often rely heavily on Siri to read text aloud. The ability to adjust pace can make Siri more understandable, especially for people who need a slower speaking rate.
Expressivity can also help with comprehension. A voice that emphasizes key words and phrases makes it easier to follow directions or understand important information. Apple has long considered accessibility features an important part of its software updates, and the expansion of these controls to more voices with Siri's neural engine is another indicator of that focus.
The new controls are not hidden behind a separate accessibility menu, but they are easy to find in the Siri settings. That is a good thing, because it means more users are likely to discover them. When features are buried, they tend to be used less often.
The Future of Siri Voice Options
It is not hard to imagine Apple expanding this feature further in future updates. The current set includes American and British English voices, but Apple may add more English accents such as Australian or Irish, or even voices in other languages. The underlying neural synthesis engine is likely flexible enough to support many accents if enough data is available.
There is also the question of whether these controls will eventually be offered on more devices. Apple's hardware requirements could change as the company optimizes its voice models. If future versions of the on-device model become smaller or more efficient, the feature could be backported to older chips.
For now, iOS 27 beta 5 makes the Siri voice customization feel like a fully realized feature. The new pace and expressivity sliders are no longer a niche add-on; they are available across the most common Siri voices. That is a promising sign for the future of Siri and a clear indication that Apple is investing in a more personalized assistant experience ahead of the Siri AI launch later this year.
Source: 9to5Mac News