How does booking a private jet work, step by step?

The process looks nothing like booking a scheduled flight – there is no fixed timetable, just an individual setup tailored to the route:

Submit a request

Provide route, date, number of passengers, luggage and any special requirements (pets, wheelchair access, a cold chain for medical equipment) – free and non-binding, via PrivatJet One.

Compare offers

Several charter operators from the PrivatJet One network submit offers with different aircraft, which are then compared by price, availability and cabin configuration.

Contract and booking confirmation

Once you choose an offer, a charter contract follows, covering flight time, price, payment terms and the operator’s details (Air Operator Certificate).

Preparation

Passport details and customs paperwork for international flights, catering preferences, luggage dimensions and – if needed – the limousine transfer are finalised.

Departure day

Arrival at the Fixed Base Operator (FBO) is usually 15–20 minutes before departure, with boarding directly at the aircraft – no classic security line to wait in.

That’s it.

PrivatJet One compares charter offers by jet class, route and price, making the otherwise opaque pricing of the business aviation market transparent.

Which aircraft types and jet classes are available?

Not every aircraft suits every route. Short distances and smaller airports call for different models than intercontinental flights:

A detailed breakdown of extra costs such as landing fees or de-icing can be found in our private jet rental experience report. Anyone wanting to compare individual models in detail can dig deeper on PrivatJet One.

What sets the classes apart?

Besides passenger count and range, three factors mainly decide which class fits a trip: cabin height (in turboprops and light jets you often have to stoop, from midsize upward you can stand upright), cruising speed (typically Mach 0.74 to 0.90), and equipment – ultra long range jets usually offer a sleeping cabin and sometimes a separate shower, which makes a real difference on a ten-hour nonstop flight.

How is the price made up: block hours and empty legs

Billing is based on block time – the time from closing the cabin door to opening it at the destination, not just the minutes actually in the air. Depending on the jet class, the hourly rate typically runs between about €2,000 and €10,000. On top of that come:

• Landing fees and handling: €800 to €2,500 per airport, sometimes considerably more at busy airports (for example during major trade fairs or sporting events).
• Positioning flight: If the aircraft is not already at the departure airport, the empty leg to get there is billed too – often at half the hourly rate.
• Crew overnight stay: On round trips with a stopover, hotel and meal costs for the crew are added.

Empty legs: cheap, but inflexible

An empty leg happens when a jet has to fly back to its home base without passengers after a charter, or needs to be repositioned for its next assignment. These legs can be booked at discounts of 50 to 75 percent compared to a regular charter price. The catch: the date, route and aircraft type are already fixed and can shift at short notice or fall through entirely – an attractive way to save money for flexible frequent flyers, but not ideal if you need to be at a fixed real estate appointment the next morning.

What charter services exist – private, business, medical and more

Charter flights aren’t just for leisure trips. Depending on the occasion, different aircraft and additional services come into play – a weekend trip for three needs a different setup than a medical repatriation with a doctor and nursing staff on board, or a delegation trip for ten people with its own security detail. Anyone looking for the right setup for their occasion will find in the table below which jet class is typical and what extra equipment or personnel can be booked alongside it:

For the medical segment, PrivatJet One offers more detail directly on privatejets.one, and the same applies to corporate travel, including how billing and tax treatment typically work.

How does the ground transfer work – limousine, driver and FBO

The comfort of chartering doesn’t end at the aircraft door. A good part of it happens on the ground:

1. FBO instead of terminal: Charter flights depart from and land at a Fixed Base Operator – a small, dedicated terminal away from scheduled traffic, with no long security or check-in lines.
2. Limousine service: A driver picks you up at home or at the office and drives directly onto the apron, right up to the aircraft – no changing between terminal and gate.
3. Handover at the destination: At the destination airport, another vehicle is usually already waiting, while the FBO ground staff take care of the luggage.
4. The last mile: For hard-to-reach destinations such as islands or ski resorts, a helicopter transfer sometimes covers the final leg.

Limousine service, FBO handling and, where needed, a helicopter connection can all be booked directly along with the request, so the result is one continuous door-to-door transfer instead of separately organised legs.

Who is this service for?

Real estate investors with nationwide appointments

Anyone who regularly views properties in several cities or buys real estate directly often saves an entire travel day with a charter flight – especially for appointments away from the major airport hubs. Investors reviewing properties as a capital investment across several regions at once can bundle multiple viewings into a single day, something scheduled flights can rarely match.

Business travellers

Short-notice meetings, several stops in one day, or destinations without good scheduled connections can be planned far more flexibly with a charter flight than with regular airline routes.

Group and delegation trips

For family offices, teams or larger groups, a Light Jet often already pays off from around four to six people – frequently costing about the same per head as a business-class ticket, without connections or waiting times.

Frequently asked questions about booking a private jet

Is the request with PrivatJet One free and non-binding?

Yes. The request via PrivatJet One is free of charge, you receive non-binding offers and only decide afterwards whether to book.

How quickly will I get an offer?

Depending on the route and availability, initial offers are usually ready within a few hours, and sometimes on the same day for short-notice requests.

Do I have to book a whole jet even if only two of us are travelling?

Yes, charter flights are billed per aircraft, not per seat. For smaller groups, a Very Light Jet is usually the better fit than a larger class.

What does an empty leg cost compared to a regular charter?

Empty legs typically cost 50 to 75 percent less, since the aircraft has to fly the route anyway. The trade-off is that the date and route are already fixed and can no longer be changed.

Is the limousine transfer included in the price?

The transfer to and from the aircraft can be added directly to the request and is billed separately – it isn’t automatically part of every offer, but it is standard with most operators in the network.

Anyone wanting to see live fleet availability can find the current overview on privatejets.one – or submit the request directly and easily right there.

The most important AI providers compared

The following table shows ten of the best-known AI providers in direct comparison. The percentage values are deliberately kept simple and describe a comprehensible assessment based on four criteria: how easily can the offering be used without prior knowledge? How privacy-friendly is it from the perspective of a German user? What is the price-to-performance ratio? And how extensive are the features (text, image, voice, video and more)?

These values are a simplified, editorial assessment for rough orientation – not a scientific measurement and not legal advice. For business use with sensitive data, the respective provider article and the current contract terms should always be reviewed.

What the four criteria mean

Ease of use

How quickly can someone without technical knowledge get the hang of it? Most major providers perform similarly well here, because the basic principle – type in a question, get an answer – barely differs between them.

Data protection/GDPR friendliness

This is where the biggest differences show up. What matters most is where a company is headquartered and where its servers are located: providers based in the European Union, such as Mistral AI, don’t need to rely on international supplementary agreements, while US providers depend on the EU-US Data Privacy Framework or Standard Contractual Clauses. For providers with servers outside the EU or the USA, for example in China, particular caution is advisable with personal or business-sensitive data.

Price-performance

How much functionality do you get for the money spent? Free or very inexpensive offerings such as Meta AI or DeepSeek naturally score well here, even if they require compromises elsewhere.

Feature scope

How broad is the offering – pure text answers, or additionally image, voice, video, and the ability to independently handle multi-step tasks?

The ten providers in brief

OpenAI – ChatGPT

The best-known name in the field of artificial intelligence. ChatGPT covers text, image, voice and video in a single product and is developed further by OpenAI at short intervals. All the details on pricing and data protection are in the provider article on OpenAI; the individual model versions are explained on the ChatGPT models page.

Anthropic – Claude

Anthropic was founded by former OpenAI employees with a particular focus on safety. Claude is considered especially reliable with long texts and documents. More on this in the provider article on Anthropic and on the Claude models page.

Google – Gemini

Gemini is deeply embedded in Gmail, Google Docs and Search, and is especially practical for anyone already living in the Google ecosystem. Details in the provider article on Google and on the Gemini models page.

Microsoft – Copilot

Copilot is built directly into Windows, Word, Excel, Outlook and Teams, making it especially strong in classic everyday office work. More in the provider article on Microsoft and on the Copilot models page.

xAI – Grok

Grok is closely integrated with the platform X, making it especially current when it comes to up-to-the-minute topics. Details in the provider article on xAI and on the Grok models page.

Meta – Meta AI and Llama

Meta AI is built directly into WhatsApp, Instagram and Facebook for free and is financed through advertising rather than a subscription. Its technical foundation, Llama, can also be self-hosted. More in the provider article on Meta and on the Llama models page.

DeepSeek

The Chinese provider impresses with very low prices and strong performance on logic and programming tasks, but is critically reviewed by German data protection authorities because of its server location in China. Details including all models (V3, R1, V4) in the article on DeepSeek.

Mistral AI

The French company is Europe’s most important own answer to the US and Chinese AI corporations and scores especially well on data protection, since servers within the EU can be chosen. More, including all models (Large, Medium, Small), in the article on Mistral AI.

Perplexity AI

Perplexity works more like an answer engine than a classic chatbot: every statement is backed up with a source citation. Details in the provider article on Perplexity and on the Perplexity models page.

Moonshot AI – Kimi

Kimi is the chatbot from the Chinese company Moonshot AI and became internationally known through the openly available model Kimi K2, which performs especially strongly on programming and agentic tasks. As with DeepSeek: with standard use, the data ends up on servers in China. Details in the provider article on Kimi and on the Kimi models page.

Which provider suits whom?

Frequently asked questions about AI models

What is the difference between an AI provider and an AI model?

The provider is the company, such as OpenAI or Google. The model is the actual technology behind it, which powers one or more products – at OpenAI, the product is called ChatGPT, while the underlying models carry their own names such as GPT-5.6.

Which AI is the best?

That can’t be answered across the board. Depending on the task, budget and data protection requirements, different providers perform differently well – the comparison table above serves as initial orientation, while the linked individual articles provide the details.

Are all these AI providers allowed in Germany?

In principle, yes – there is no general ban on any of the providers named. However, for individual providers with server locations outside the EU, particular caution is advised with personal or business-sensitive data – details on this are in the respective provider article.

Do I always have to pay for good results?

No. Several providers offer usable free versions, such as Meta AI, DeepSeek, Kimi, or the basic versions of ChatGPT, Gemini and Copilot. For regular or more demanding use, an inexpensive entry-level subscription is usually worthwhile.

What does this have to do with Lukinski?

As a real estate company, we ourselves work intensively with artificial intelligence – for example with Lukinski AI, our own AI assistant for real estate questions, hosted in Germany. This cluster additionally puts the large, general-purpose AI providers into context, so that you can make an informed decision independent of our own offering.

From research model to heavyweight: the Kimi K2 series

Kimi K2: the technical foundation

The original Kimi K2 version was introduced in July 2025 and described in detail in a technical report (“Kimi K2: Open Agentic Intelligence”). Technically, it is a so-called Mixture-of-Experts (MoE) model: instead of a single, continuously active neural network, the model consists of many specialized sub-networks (“experts”), of which only a small portion is actually activated per request. For Kimi K2, according to Moonshot AI, there are a total of around one trillion parameters, of which around 32 billion are actively used for any single request. The model is spread across 384 experts, of which 8 are specifically selected per text segment, supplemented by an always-active “shared” expert.

This structure explains why Kimi K2, despite its enormous overall size, can be operated comparatively economically: the entire trillion parameters do not need to be computed for every request, only a small, selected part. According to Moonshot AI, the model was trained using a specially developed optimization method (MuonClip), intended to improve training stability for such large MoE models. The original base version offered a context window of 128,000 tokens; the Instruct version (for dialogue and instructions) released at the same time was expanded to 256,000 tokens.

Kimi K2 Thinking: thinking and acting in one model

In autumn 2025, according to consistent reports from November 2025, Moonshot AI introduced “Kimi K2 Thinking,” a variant that additionally incorporates extensive, multi-step thinking before the actual answer and, while still in the thinking process, can already call on tools (such as a web search or a calculator). According to available benchmark figures, the model achieved a score of around 71 percent on the well-known programming test “SWE-Bench Verified” and, in test scenarios, was able to independently carry out several hundred consecutive tool calls to solve complex, lengthy tasks. This deliberately positioned Moonshot AI in the field of so-called agent models, which don’t just answer but independently plan and execute multi-step tasks.

Further development in 2026: K2.5, K2.6 and K3

Over the course of 2026, Moonshot AI released several further tiers. According to consistent, though not always uniform, sources, K2.5 appeared, followed in April 2026 by K2.6 with a context window expanded to 256,000 tokens and improved programming capabilities, followed by a variant specialized in programming tasks called K2.7 Code. Finally, in July 2026, Kimi K3 followed, which, according to the manufacturer, comes with a context window enlarged to around 1 million tokens and native image processing (vision) – according to Moonshot AI, four times the size of K2.6’s context window. Various sources give differing figures for K3’s exact parameter size (some in the range of roughly two to three trillion total parameters); a single, officially definitive figure cannot currently be confirmed with complete certainty, which is why we deliberately speak of an order of magnitude rather than an exact number here.

What makes the Kimi models technically and strategically distinctive?

Open weights with a commercial condition

Moonshot AI releases the model weights for K2 and its successors under a self-formulated, modified MIT license. At its core, this is a very permissive open-source license that also allows commercial use. One special feature: anyone operating a product or service based on the models with more than 100 million monthly active users or more than 20 million US dollars in monthly revenue must visibly display the notice “Kimi K2” in their own interface. For the vast majority of smaller and medium-sized use cases – including in the real estate industry – this clause practically never applies, but it shows that “open” at Kimi is not meant entirely unconditionally.

Agentic capabilities and a large context window

The strategic focus of the Kimi models is clearly on so-called agentic capabilities: the ability to independently handle multi-step tasks with tool use, web access and code execution, rather than just answering individual questions. Combined with a very large context window – around one million tokens for K3 – this makes it possible to process very extensive document collections in a single pass, such as several hundred pages of contract documents or a larger collection of property documents.

Rule of thumb: the large context window makes it possible to read in a great many documents at once – but this does not replace careful review of the output, since even a large context window does not protect against errors or omissions in the result.

Practical examples of use

1. Automated summarizing of large file collections

A case worker at a property management company could have several hundred utility bill statements or minutes from owners’ meetings read in via the API in a single pass, in order to automatically flag anomalies – such as unusually high cost items. The large context window of K2.6 or K3 makes this technically possible without first having to split the documents into many small chunks.

2. Drafting and laying out a listing exposé

In agent mode, a real estate agent can have a first, multi-page draft version of a property exposé created, including a simple web presentation, which is then reviewed editorially by hand and cross-checked against the actual, legally binding property data.

3. Programming support for internal tools

Since the Kimi models perform particularly strongly on programming benchmarks, some IT departments at real estate companies use Kimi to help build small internal automation scripts, for example to reconcile rent lists between two software systems.

4. Research and market monitoring

Via the web-browsing feature of agent mode, a model can be used to continuously monitor publicly accessible market data, for example to summarize interest rate trends or publicly accessible real estate market reports – it’s important here to work exclusively with publicly available, non-personal sources.

5. Multilingual communication

For international real estate portfolios, a financial case worker can use Kimi to translate standard letters or summaries between several languages – though the restriction on personal content described above still applies here as well.

Rule of thumb: the larger the volume of data a model processes in one pass, the more important the question becomes of exactly where that data resides during processing – performance and data protection are two separate questions, and both need to be answered.

FAQ about the Kimi models

What is the difference between Kimi K2 and Kimi K2 Thinking?

Kimi K2 is the original base model; Kimi K2 Thinking is a variant introduced in November 2025 that incorporates more extensive, multi-step thinking with tool use before answering, and thereby performs better on complex tasks.

How large is the context window of the current models?

The models in the K2 series generally offer 256,000 tokens of context; Kimi K3, introduced in July 2026, is stated to have around one million tokens.

Are the Kimi models really open source?

The model weights are openly accessible and released under a modified, very permissive MIT license. One special feature is the attribution requirement for very large commercial applications (over 100 million users or 20 million US dollars in monthly revenue).

Are the models suited to programming tasks?

Yes, the Kimi K2 series, and especially the specialized K2.7 Code variant, perform comparatively strongly on public programming benchmarks and are frequently used for coding agents.

Can I run the models myself instead of using the Kimi website?

Yes, since the weights are released openly, they can be run via various cloud providers or on your own hardware. However, this requires technical expertise and sufficient computing capacity, since these are very large models.

Basic information about the company Moonshot AI, pricing and data protection can be found on our provider page. A comparison of all ten AI providers in the cluster is available on the overview page.

Who is behind Kimi? The history of Moonshot AI

• Kimi website
• Moonshot AI website

Moonshot AI was founded in Beijing in March 2023. Among its founders is Yang Zhilin, an AI researcher with a doctorate from Carnegie Mellon University who had previously co-authored well-known language model papers such as XLNet and Transformer-XL. Together with other co-founders, he built a company that Chinese media often mention in the same breath as ByteDance and OpenAI – as an example of a particularly ambitious AI startup.

Financially, Moonshot AI has been backed by, among others, the technology corporations Alibaba and Tencent as well as investors such as IDG Capital. Its company valuation rose considerably within a short time: from around 300 million US dollars at founding, to about 2.5 billion US dollars in early 2024, up to figures in the multi-billion US dollar range reported for early 2026. This makes Moonshot AI one of the most highly valued private AI companies in China.

From niche chatbot to internationally recognized model

Kimi initially became known as a Chinese-language chatbot. Its international breakthrough came in summer 2025 with the Kimi K2 model series, introduced as an open model with a very large parameter count, which caused a stir in expert circles because it could compete with considerably more expensive models from Western providers on benchmarks for programming tasks and agentic capabilities.

Kimi as a product: chat app, agent mode and API

For most users, Kimi is first and foremost the web and app interface at kimi.com. There, the model can be used like an ordinary chatbot for texts, translations, research or programming tasks. Since autumn 2025, Moonshot AI has additionally offered an extended agent mode, described in reports under names such as “OK Computer” or “Kimi Work.”

Agent mode

In this mode, Kimi can independently take on multi-step tasks: it can research on the web, create spreadsheets and simple multi-page websites, run Python code, and, according to Moonshot AI, read in several dozen files at once. For a case worker in property management, this could theoretically mean that a larger amount of documents – such as several utility bill statements or listing exposé drafts – is summarized in one pass. Important here: all uploaded content is processed on Moonshot AI’s servers, more on that below.

Kimi via the API

Developers and companies can also access the Kimi models via their own programming interface (API), for example to integrate them into existing software such as a CRM system or document management. This interface is offered directly by Moonshot AI as well as by several third-party providers, through which the open model weights can also be run on your own or European cloud infrastructure. This is the decisive difference from pure chat use: anyone who self-hosts the open weights also decides for themselves where the data is stored.

Rule of thumb: Kimi is not “one” product, but two – the convenient chat and agent access with data processing in China, and the openly available model weights, which can in principle also run on your own infrastructure.

Pricing models: what does Kimi cost?

For private use via the chat interface, Moonshot AI offers free basic use with restrictions on request volume and feature scope, supplemented by paid packages for more capacity. For developers and businesses, API billing by text volume (tokens) is the most relevant. The following figures reflect the publicly available price lists as of August 2026 – as with all AI providers, these prices can change on short notice, so it’s always worth checking Moonshot AI’s current pricing page before doing a business calculation.

Practical example: a property management company has 500 utility bill statements automatically summarized.

• 1,500 words per statement
• Around 750,000 words in total
• About 1 to 1.3 million input tokens
• K2 series: under 2 US dollars for input processing
• K3: correspondingly higher costs

On top of the input comes the usually more expensive output – the summary created.

Data protection and GDPR: what does the server location China mean?

For use at a German business, this point is the most important one: if you use Kimi via the website kimi.com or via Moonshot AI’s official API, your inputs are processed on servers in China. China currently has no so-called adequacy decision from the European Commission – this means the EU does not officially classify the level of data protection there as equivalent to the General Data Protection Regulation (GDPR). Similar concerns are also discussed with other Chinese AI providers such as DeepSeek.

According to Moonshot AI’s publicly available terms of use, entered content can also be used to further develop its own models. For purely private, non-binding use, that may be an acceptable risk. For a case worker who wants to process personal data from a lease, a credit report, or a customer inquiry, however, the starting position is different: here, the strict GDPR requirements for data processing apply, and transferring such data to a provider headquartered and running servers in China can hardly be made legally sound without additional contractual and technical safeguards.

Rule of thumb: anyone using Kimi for general research or non-binding text drafts without personal data involved takes on a different risk than someone entering tenant data, names, or contract content. The latter should be avoided without additional safeguards.

A conceivable alternative for privacy-sensitive use cases is running the open Kimi model weights on your own or European server infrastructure, so that the data never has to leave the country – however, this requires your own technical expertise and corresponding hardware or cloud costs. For businesses in the real estate industry looking for an AI solution with German hosting and without this configuration effort, specialized offerings such as Lukinski AI are available as an alternative, designed from the outset for German server locations and the industry’s requirements.

Strengths and weaknesses of Kimi

Who is Kimi suited for?

Kimi is particularly well suited to certain groups of users.

• Technically skilled users and developers
• Companies with their own hosting
• Programming and automation tasks

For real estate agents with listing exposé texts or property managers with correspondence, caution is advised as soon as personal data is involved. Here, a comparison with European providers is worthwhile.

Frequently asked questions about Kimi and Moonshot AI (FAQ)

Can Kimi be used for free?

Yes, there is free basic use with restrictions via the chat interface. For higher usage volumes or API integration, costs apply based on the amount of text processed.

Is Kimi safe for business or personal data?

When using Moonshot AI’s standard chat or API access, data is processed in China, for which there is no EU adequacy decision. Special caution is therefore advised for personal or contractually sensitive data from the real estate or finance sector.

What distinguishes Kimi from ChatGPT, Claude or DeepSeek?

Kimi differs from most Western providers mainly through the open release of its model weights. There are clear parallels with DeepSeek, another Chinese provider, regarding data protection issues and server location.

Can Kimi also be run without Chinese servers?

In principle yes, since the base models are released as open weights and can also be made available on European infrastructure via various third-party providers. However, this requires technical expertise and your own server capacity.

What does the model name K2 stand for?

K2 refers to the large model generation from Moonshot AI introduced in summer 2025, named after the second-highest mountain on Earth. Details on the individual K2 versions and their successors can be found on the in-depth models page.

More technical details on the individual model versions can be found on our page Kimi models in detail. A comparison of all ten AI providers covered in the cluster is available on the overview page.

The Grok model family at a glance

Grok 3 – the previous generation

Grok 3 was xAI’s main model for a long time and now forms the basis for the free tier. It handles the basic tasks of a chatbot – answering questions, drafting texts, simple research – but offers a considerably smaller context window and less accuracy for complex, multi-step tasks than the newer models.

Grok 4 and Grok 4 Heavy

Grok 4 was the next big step and brought considerably better results in logical reasoning and more complex questions. The Grok 4 Heavy variant goes a step further: for difficult tasks, it uses several parallel thinking processes and compares the results before giving an answer. This makes Grok 4 Heavy slower and considerably more expensive, but also more reliable for tricky tasks – such as mathematical calculations or multi-step analyses.

Grok 4.1 Fast – fast and inexpensive

For tasks where speed and low cost matter most – such as automatically answering many short inquiries in an app – xAI offers Grok 4.1 Fast, a stripped-down but very fast and inexpensive variant. Its depth of content is lower, but that is usually sufficient for simple, repetitive tasks.

Grok 4.20 – a huge context window and multi-agent technology

With Grok 4.20, xAI introduced a model in early 2026 that internally works with several specialized “roles” that cross-check each other before an answer is given – put simply: one role coordinates, one checks facts, one assesses technical details, one provides creative phrasing. For users, this mainly shows up as more reliable answers for longer, more complicated queries. In addition, there is a very large context window of around 2 million text units (tokens) – equivalent to several thousand pages of text that the model can keep “in mind” at once, for example for analyzing longer contracts or extensive data collections.

Grok 4.5 – the current focus on complex and technical tasks

Grok 4.5, released in July 2026, is currently xAI’s most demanding available model. It was specifically trained for longer, demanding work sessions – originally mainly for software development, but now also for other complex, multi-step tasks such as comprehensive research or the analysis of larger data sets. Its context window is around 500,000 tokens, so smaller than Grok 4.20’s, but the model works more precisely on difficult reasoning tasks. Via the programming interface, Grok 4.5 costs around 2 euros per million input text units and around 6 euros per million output text units.

Grok 5 – what is known so far

A successor model with the working name Grok 5 is, according to xAI, in the training phase on an expanded Colossus data center. There is no official release date as of August 2026; according to rumors, the model is expected to be considerably larger than Grok 4 and to be introduced later in 2026. Until the official announcement, Grok 4.5 remains the most powerful generally available model.

Rule of thumb: the higher the version number, the more powerful and usually also more expensive the model – but not every task needs the strongest model. For simple everyday questions, the fast, inexpensive variant is often enough.

Comparison table of the Grok versions

Real-time data access via X

One feature shared by all current Grok models is direct access to public posts on the platform X. This allows Grok to react to events that are only a few minutes old – a clear difference from models whose knowledge is limited to a certain training cutoff date. The downside: posts on social media are not automatically verified or reliable, which is why answers based on current X posts should always be read with a healthy dose of caution, especially for controversial or unclear topics.

Image and video generation with Aurora and Grok Imagine

For generating images, xAI uses its own system called Aurora, now marketed under the product name Grok Imagine. It generates images from text descriptions and, by now, also short videos a few seconds long in various formats. In the paid SuperGrok tier, the full feature set of Grok Imagine is included; the free tier offers only a limited basic version. Via the programming interface, video generation is billed by the number of seconds generated, which makes it comparatively inexpensive for short clips.

Examples of use from everyday life and business

Example 1 – everyday life: A user is planning a trip and has Grok summarize, based on current posts on X, what the weather and traffic conditions currently look like at the destination – something a model without real-time access cannot deliver.

Example 2 – small business: A case worker at a small trades business uses Grok 4.1 Fast to pre-sort incoming customer inquiries by email and draft reply suggestions, which are then reviewed by a human.

Example 3 – real estate and finance: A real estate office wants to get a first impression of the current sentiment around a particular location – for example, whether there has recently been more discussion about rising rents or a new construction project in a city. Thanks to its access to current X posts, Grok can summarize such discussions and provide initial clues. Importantly: such a sentiment analysis does not replace a robust market analysis based on official data, but merely provides a quick, supplementary impression.

Example 4 – reviewing a long document: A case worker uploads a longer lease or an extensive tender document into Grok 4.20 and has the key clauses, deadlines and figures summarized – thanks to the large context window, the model can take the entire text into account at once, instead of having to break it into sections.

Example 5 – a technical task: A developer uses Grok 4.5 to analyze an existing software application across several files and get suggestions for improving the program code – a task the model was specifically trained for.

Rule of thumb: the longer and more branched the task, the more important the context window becomes – for short everyday questions, on the other hand, it hardly matters.

Practical example: estimating the cost of the programming interface

A real estate office wants to have around 50 property descriptions automatically drafted per month with Grok 4.5. Each description requires an estimated 1,000 tokens of input (property data, conditions) and 1,500 tokens of output (finished text).

• Input: 50 x 1,000 = 50,000 tokens, at 2 euros per million tokens that comes to around 0.10 euros
• Output: 50 x 1,500 = 75,000 tokens, at 6 euros per million tokens that comes to around 0.45 euros
• Total cost per month: around 0.55 euros

This calculation example shows: the pure text costs via the programming interface are very low for manageable text volumes – the actual costs usually come from developing and maintaining the integration, not from the individual requests themselves.

Frequently asked questions (FAQ)

Which Grok model should I use as a beginner?

For getting started, the standard model in the free or the cheapest paid tier is generally sufficient. Switching to more powerful models is only worthwhile once you hit limits with complex or long tasks.

What does “context window” mean in simple terms?

The context window describes how much text a model can keep “in mind” at once. A large context window is important if you want to have long documents analyzed at once, but it hardly matters for short questions.

Is Grok 4.20 the same as Grok 4?

No. Grok 4.20 is its own, later model generation with a considerably larger context window and an internal multi-role check that was not yet present in Grok 4.

When is Grok 5 coming?

There is no official date as of August 2026. xAI has confirmed that a successor model is in the works, without naming an exact date.

Can I get reliable market data for real estate decisions with Grok?

Grok can summarize current discussions and sentiment, but it does not replace verified official data or a professional market analysis. It is best suited as a quick supplement, not as the sole basis for a decision.

An overview of the company xAI itself, its pricing tiers and the data protection assessment is available on our xAI provider page. Anyone wanting to see the other major AI providers compared instead will find the overview on the AI models comparison page.