AI's Big Five: OpenAI, Anthropic, Google, Microsoft and GrokAI's Big Five: OpenAI,Anthropic, Google,Microsoft and Grok

That matters in Kenya because the choice is becoming a practical business decision. A student may want affordable research help, a developer may need reliable code assistance, while a company may be deciding whether to pay for AI tools for dozens or hundreds of employees. The best service is not automatically the one with the highest benchmark score. It is the one that performs the required work accurately, consistently and at a price the user can sustain.

There is no single, authoritative market-share table that ranks all five across every part of the AI industry. Some studies measure chatbot traffic, others track business subscriptions, model revenue, cloud contracts or software purchases. Those measures tell different stories and should not be treated as interchangeable.

OpenAI: the broad consumer platform

OpenAI has built one of the best-known AI brands through ChatGPT, while expanding into image generation, voice, research, coding and business automation. Its advantage is breadth: a single subscription can support writing, analysis, data work, software development and creative tasks. Its developer API and enterprise products also give companies ways to build AI into their own services.

The weakness of a broad platform is that breadth does not guarantee the best result on every specialist task. Usage limits, model selection and the cost of higher tiers also matter. OpenAI's ChatGPT plans range from free access to paid individual and business offerings. Its paid tiers offer more advanced reasoning, research, coding and productivity features, but the available limits vary by plan.

Anthropic: the professional and coding challenger

Anthropic has positioned Claude around careful reasoning, long documents, software development and business use. Claude Code has become an important part of its offer to developers, while Projects, Research and integrations make the service useful for people working with large amounts of information.

Its commercial momentum is striking. Cledara's software purchasing data for the 12 months to 31 October 2026 put Anthropic and Claude first in its tracked software purchases, accounting for 12.3% of all purchases and renewals in the dataset. OpenAI and ChatGPT accounted for 5.9%. This is payment activity among Cledara's customers, not global market share or the percentage of all businesses using each service.

Anthropic's published pricing lists Claude Pro at US$20 (about KSh2,600) per month when billed monthly, with an annual-billing discount. Its Max plans start at US$100 (about KSh13,000) per month and offer higher usage limits. Team pricing starts at US$25 (about KSh3,250) per person per month with annual billing, with a five-seat minimum. The company says usage limits still apply.

Google: distribution, search and an existing ecosystem

Google's strength is not just Gemini's model quality. It can put AI inside products people already use, including Search, Gmail, Docs, Sheets, Android and Google Cloud. That gives it a route to millions of users without requiring each person to adopt a new standalone chatbot. Its large cloud business and access to its own computing infrastructure are further advantages.

Google's AI subscriptions bundle Gemini access with storage and, depending on the tier and country, creative tools and features across its services. Its Google AI plans include Pro and Ultra tiers.

The company lists Ultra from US$99.99 (about KSh13,000) a month in the United States, with higher usage limits and additional features. Plan names, availability and prices can vary by market, so Kenyan customers should check the local checkout price before subscribing.

Google's challenge is converting its huge reach into regular paid use. A person who gets useful AI features through an existing Google account may not see a reason to pay for a separate premium plan.

Microsoft: the workplace and cloud distribution giant

Microsoft is different from the other names on this list. It is not simply selling one consumer chatbot or relying on one frontier model. It distributes AI through Microsoft 365 Copilot, GitHub Copilot, Windows, Teams and Azure, and offers businesses access to models through its cloud platform.

That gives Microsoft a powerful advantage in organisations already using Word, Excel, Outlook, Teams and other Microsoft products. AI can be introduced inside familiar work tools, with administration, security and compliance features that matter to large employers.

Microsoft lists Copilot Business at promotional prices from US$18 (about KSh2,340) per user per month with annual billing on its business pricing page. Other bundles and monthly payment options cost more. This is a business product, not a directly comparable consumer chatbot subscription. The price of the wider Microsoft 365 licence may also need to be considered.

Its main challenge is proving that AI features save enough time or money to justify another recurring software bill. Having access to a model is not the same as getting a measurable improvement in productivity.

xAI: real-time information and the X connection

xAI's Grok competes through its integration with X, giving it a route to current public conversation and real-time social information. It also offers web search, voice, image and video generation, and developer and enterprise services.

That access can be useful when users want to follow fast-moving online discussions. But information that is current is not automatically accurate. Posts can be misleading, unverified or deliberately manipulated, so Grok's real-time reach does not remove the need to check original sources.

xAI lists SuperGrok at US$30 (about KSh3,900) per month and SuperGrok Plus at US$100 (about KSh13,000) per month on its pricing page. Higher tiers offer increased usage and additional features. As with the other services, availability and final prices may differ by country.

Who is actually winning?

The answer changes depending on what is being measured.

Cledara's October payment-activity tracker puts Anthropic ahead of OpenAI among software purchases and renewals made by its customers. Its figures show a clear willingness among that business sample to pay for Claude. They do not establish that Claude has more users worldwide.

A separate Counterpoint Research snapshot for the first quarter of 2026 estimated that Anthropic held 31.4% of global large language model revenue, compared with 29% for OpenAI, 12.1% for Google and 7.2% for Microsoft. The study estimated 3.8 billion monthly active users across the services it tracked.

These are research estimates, not audited company accounts, and the figures depend on the study's definitions and coverage.

On model quality, the gap is also narrowing. Stanford University's 2026 AI Index technical performance chapter said that, as of March 2026, Anthropic, xAI, Google and OpenAI were clustered within 25 points on the Arena Elo ratings it cited. The finding suggests that no single provider has an overwhelming lead across all tasks.

A newer Artificial Analysis Intelligence Index update published on 27 October 2026 placed Claude Opus 5.5 at the top, followed closely by OpenAI's GPT-6 Astra and GPT-6.1 Sol, with Google's Gemini 4 Argon next. That is one composite evaluation, weighted across knowledge work, coding, science and general tasks. It is not a universal verdict on which chatbot is best for every user.

Subscription prices are only part of the cost

The entry price can be misleading. Providers limit how much heavy work subscribers can do, and the limits may vary by model, task and demand. A plan that looks cheap can become expensive if a user repeatedly hits usage caps and needs a higher tier.

Business buyers also need to consider the number of seats, administrator controls, data protection terms, integrations, support, API charges and the cost of checking AI-generated work. API usage is generally billed separately from consumer subscriptions. A company building an AI-powered product therefore needs to compare the cost per completed task, not only the monthly chatbot fee.

For Kenyan customers, foreign-currency billing adds another variable. At an indicative rate of KSh130 to the US dollar, a US$20 subscription is about KSh2,600 before applicable taxes, bank charges or exchange-rate differences. The actual amount charged can be higher or lower.

Quality means more than a clever answer

The leading services are increasingly close on some tests, but their strengths differ by task and change as providers release new models. Claude is a strong option to test for long documents and coding workflows. ChatGPT offers a broad set of tools in one service.

Gemini benefits from Google's ecosystem and its integration with existing products. Copilot is attractive where a business already relies on Microsoft software. Grok's X connection can help with current online discussion, although its results still require verification.

None should be treated as an authority simply because it sounds confident. All can make up facts, misunderstand instructions, mishandle figures or cite material that does not support an answer. Performance also depends on the model selected, the prompt, the available tools and whether the task requires current information.

For Kenyan users, local usefulness deserves its own test. Can the service understand Kenyan English and Kiswahili? Does it handle KSh figures, local tax rules and Kenyan institutions accurately? Can it work with the documents and applications a business already uses? Does it protect confidential customer, employee and financial information? These questions may matter more than a small difference in a global benchmark.

The business model is shifting from chat to work

The contest is moving beyond monthly access to a chatbot. Providers want customers to use AI for coding, research, customer service, office work, data analysis and automated tasks. That creates more opportunities to charge businesses, but it also raises the bar: companies will expect measurable gains rather than impressive demonstrations.

Mwenendo's earlier analysis of the big technology companies' AI spending race examined the huge investments behind the competition. The next test is whether those investments produce services customers keep paying for, at prices that allow providers to make money.

For consumers and businesses, the sensible approach is not to subscribe to every leading service. Identify the work that needs to be done, test two or three providers on the same real tasks, check the accuracy of their outputs, measure the time saved and compare the full cost. The winner is the tool that delivers dependable results for the job, not necessarily the company with the loudest launch or the highest score on one leaderboard.