Alfred

What is Alfred?

Alfred was originally a GPT-based chat assistant designed for iPhones, developed by the Dreamsands team. At the beginning of 2023 it was renamed Frank, and it gradually acquired features such as online search, image generation, an AI keyboard, voice dialogue, and a community for prompt creation.

As of this verification, the same app store identifier is shown as AI-Story; however, the page descriptions still use the name Frank. It is not the Mac productivity tool Alfred, nor is it any other customer service robot with the same name – it is necessary to identify it based on its developer and app identifier before using it.

Name and version history

Core functions

• Engage in continuous dialogue with AI through text.
• Use voice to ask questions and listen to voice responses.
• Search for more recent public information.
• Generate text and images based on the prompts.
• Identify the content of images and perform multimodal question answering.
• Use an AI keyboard to assist with writing in other applications.
• Browse, save, and test community prompts.
• Modify or rewrite the existing answer based on the context.

Text chat

Users can ask questions, write, summarize, translate, or brainstorm in natural language, and they can refine the results by asking follow-up questions. The responses are generated by third-party large models; however, complex facts, numbers, and professional advice still require independent verification.

Voice dialogue

The application supports voice input and voice output; future versions will offer an even more seamless real-time conversation experience. The voice feature is useful for commuting, practicing speech, and operating the device without using hands, but it is important to be mindful of privacy and recording permissions in public settings.

Online search

The Frank phase introduced search and browsing capabilities without traditional ads, which can be used to look up information about restaurants, events, prices, and general details. Whether the search function is real-time, whether the information provided is complete, and the availability in different regions depend on the results shown within the application.

Text generation

• Draft emails, messages, and social media content.
• Rewrite the tone and condense long texts.
• Generate an article outline and topic directions.
• Translate and explain unfamiliar expressions.
• Organize meeting questions and to-do lists.
• Generate practice questions for the learning topic.
• Multiple versions of the existing copy are provided.

Image generation

Users can create images by providing a textual description, and they can request adjustments to the theme, visuals, and style. The resulting content may contain errors in terms of character structure, text, and brand elements; therefore, it is necessary to verify portraits, trademarks, and licensing for use in commercial purposes.

Image understanding

The specifications for the GPT-4o versions emphasize multi-modal processing of audio, visual, and text data. Users can upload images and ask questions about their content, but IDs, medical images, and other sensitive materials should not be uploaded when it is unclear how such data will be processed.

Frank AI keyboard

Keyboard extensions bring the AI writing functionality to email, social media platforms, and other applications that support keyboard input. They can be used to run custom prompts, continue writing text, or generate images; however, it is important to carefully examine the full access permissions before enabling third-party keyboards.

• Quickly rewrite text in different applications.
• Save frequently used custom prompts.
• Create a personal quick action menu.
• Generate short posts, emails, and replies.
• Reduce the need to switch back and forth between chat apps.

Prompt community

During the Frank phase, an open prompt community is available; users can browse categories, view popular prompts, test them out on the spot, and import keyboard shortcuts. The quality of content in this community varies, and prompts that involve unauthorized requests, false information, or unsafe elements should not be used.

Memory capacity

Earlier versions of Alfred allowed users to enter personal information that they wanted the assistant to remember during conversations. This memory feature helps avoid repeating information, but passwords, identification numbers, financial data, or any other sensitive information that cannot be revoked should not be stored as long-term memories.

Supported models

Earlier versions used GPT-3, and later capabilities related to ChatGPT, GPT-4, and GPT-4o were added. The model names, quotas, and actual routing on this page may change; therefore, it is not possible to judge the current services based solely on the previous update records.

Supported platforms

Which users are it suitable for

• Users who wish to use a universal AI assistant on the iPhone.
• People who need to make voice queries and receive voice responses.
• Users who frequently write content in multiple applications.
• Professionals who need to quickly rewrite emails and messages.
• Beginners who want to browse and save prompts.
• Individuals who need simple image generation and multimodal question answering.
• Users who are willing to verify AI-generated answers and manage subscription costs.

Typical use cases

• I can answer general knowledge questions at any time.
• Generate drafts for emails and social posts.
• Use voice for brainstorming.
• Translate short texts related to travel or work.
• Search for restaurants, competitions, and public information.
• Compress long content into key points.
• Create simple illustrations and creative concepts.
• Quickly rewrite replies using the system keyboard.

It’s not very suitable for which situations

• Teams that need to stabilize enterprise SLAs and ensure administrator control.
• Developers who require that the models, data, and processing pipelines be entirely open source.
• Users who need to enjoy a full native landscape experience on the iPad.
• Those who wish to use the original Alfred brand and the old functional interface.
• Users who cannot accept frequent changes in subscription prices and product names.
• Used to handle high-risk decision-making in areas such as medicine, law, and investing.
• Enterprises that require clear API integration and batch automation.

Start using the tutorial

1. Check the developer and app numbers in the app store.
2. Verify the current name, version, and system compatibility.
3. After downloading, read the privacy policy and subscription terms.
4. First, use non-sensitive questions to test text chatting.
5. Enable microphone, camera, or keyboard permissions as needed.
6. Check the free quota and auto-renewal cycle before making a purchase.
7. Important answers are verified using other reliable sources.

Tutorial on using an AI keyboard

1. Initialize keyboard expansion within the application.
2. Enter the system keyboard settings to add the corresponding keyboard.
3. Check whether full access really needs to be enabled.
4. Switch to the AI keyboard in the non-sensitive text field.
5. Choose to rewrite, generate, or customize the prompt.
6. Confirm the output content before sending.
7. Permissions can be revoked from the system settings when not in use.

Voice dialogue tutorial

1. Allow the app to access the microphone while it is in use.
2. Choose between text, normal voice, or continuous conversation mode.
3. Start the conversation with a complete question.
4. Check whether the recognized text is accurate.
5. By asking follow-up questions, the length, tone, or output format can be specified.
6. Stop voice input when personal information is involved.

Ways to improve the quality of responses

• First, clarify the goal, target audience, and desired format.
• Provide the necessary and non-sensitive background for the task.
• Break down complex tasks into multiple steps.
• It is necessary to distinguish between facts, assumptions, and suggestions.
• Check the dates of the online search results against the original sources.
• Manually edit the draft instead of publishing it directly.

Price and subscription

The app can be downloaded for free, but the advanced features are available through in-app purchases. The page for the United States lists multiple past or current purchases; the names and amounts of these purchases may overlap, and the final cost should be as indicated on the payment page of the device.

Subscription notes

• Before purchasing, confirm whether the price is for a week, month, year, or as a one-time fee.
• Check the amount of automatic renewal after the free trial ends.
• Projects with the same name may correspond to different historical versions.
• Canceling the subscription does not result in an automatic refund of the amounts already paid.
• Regional taxes and currencies can affect the actual price.
• The permanent version does not necessarily guarantee unlimited use of all future models.
• Save the purchase receipt and check the subscription status regularly.

Privacy and data

The privacy disclosures in app stores indicate that developers may handle contact information, user content, search history, usage data, and diagnostic data. Some of this data may be linked to the user’s identity, and the specifics vary depending on the features, region, and version.

• Do not submit account passwords or payment information.
• Avoid uploading identification documents or sensitive images of others.
• Pay attention to the conversations around you before turning on the microphone.
• When using keyboard extensions, be mindful of full access permissions.
• Regularly delete unnecessary chats and memories.
• Use tools approved by the company in jobs that require high levels of privacy.

Accuracy and real-time performance

Large models may generate answers that sound reasonable but are not accurate; the mention of a particular model name does not guarantee that the information is up-to-date. Search results, dates, citations, and specific figures should all be verified by consulting the original sources.

Current status of the product

The name displayed on the current app store page is AI-Story; the description still refers to Frank as the assistant. The latest version recorded dates back to 2024. The original name Alfred appears only in the historical update notes, so it should be marked in the catalog as a product that has been renamed.

Explanation of fields on the official website

The original website of the database pointed to an old designer site that had nothing to do with AI assistants, and therefore could not serve as a basis for this product. To identify this tool, one should rely on the developer Dreamsands, the application identifier, and the version history from Alfred to Frank.

API and open-source status

No official API or complete source code repository available for developers has been found for this mobile application. It is a closed-source consumer application; the use of third-party models in its underlying operations does not mean that the application itself is open source.

Product advantages

• Combine text, voice, vision, and image generation in a mobile application.
• It supports continuous follow-up questions and content rewriting.
• The AI keyboard allows writing capabilities to be utilized in other applications.
• The prompt community lowers the barriers to entry for new users.
• It supports product localization in over a hundred languages.
• The system requirements for the iPhone are relatively low.
• Various purchase options are available, including weekly, annual, and one-time purchases.

Usage restrictions

• The product has had its name changed several times, resulting in inconsistencies between the brand name and the information on the page.
• The fields on the database’s original website are no longer valid and point to unrelated websites.
• There are multiple overlapping subscription options in the app store.
• The history of the public version is quite old, so caution is needed when assessing its current level of maintenance.
• Designed solely for iPhone; there are no native iPad layout instructions.
• AI answers may be incorrect or outdated.
• Keyboard, voice, and image functions involve significant privacy permissions.
• The developer API and the complete source code are not available publicly.

Basic information

Recommendation score

The comprehensive recommendation score is 3.2 out of 5 points. It was among the first to introduce GPT chat, search, and an AI keyboard to mobile devices; however, it has undergone multiple name changes, its pricing options are complex, and its maintenance status is not clear. Before downloading, it is important to check the current version and the subscription details.

Frequently Asked Questions

What is Alfred’s name now?

It was renamed Frank in 2023; the store page with the same application identifier currently shows as AI-Story.

Are Alfred and Mac Efficiency Tools the same product?

No, this entry refers to the mobile AI chat application launched by Dreamsands.

Can Alfred be used for free?

It can be downloaded for free, but advanced models, unlimited usage, or other benefits may require in-app purchases.

Is voice chat supported?

It supports voice input and voice responses; future versions will also offer a more seamless conversational experience.

Is it possible to generate images?

It supports generating images based on prompts, and offers a creation process that integrates with text-based conversations.

Is it compatible with the iPad?

The page is marked as being designed solely for iPhone; the interface on iPad may not feature a native wide-screen layout.

How much is Alfred?

In the United States, the options available are a weekly fee of 4.99 or 6.99 dollars, an annual fee of 144.99 or 179.99 dollars, and a permanent subscription at 199.99 dollars.

Why can’t the official website be accessed?

The field in the original website’s catalog points to an unrelated old site; products should be identified using the developer and application IDs available in the app store.

Is Alfred open source?

No official, complete source code repository was found, nor are there any public APIs available for developers.