Which AI model is best? GPT, Claude, Gemini and the rest

No single AI model is best at everything, and the order at the top changes every few months. For everyday drafting, summarising and extracting, the top few models are close enough that a user would not notice the difference. They differ on harder work, such as very long documents, difficult reasoning, code, images and audio, and the best one for a team is the one that does best on its own work.

A handful of labs build the models that nearly every AI product in a business runs on, including the ones with somebody else’s logo on them: ChatGPT is a product wrapped around OpenAI’s models, and the layer underneath is usually somebody else’s.

Each lab has a model at the top of its range: the newest, the largest, the most expensive to run, the one the lab’s reputation gets measured on and the one the announcements are about. The shorthand for that group, across all the labs, is the frontier. Most work never touches it.

Who makes the models

The useful split is how the model reaches a business. One group keeps the model on its own machines and sells access to it, through an API for whoever is building something, or through a product anyone can open.

GPT, from OpenAI. The name most people outside the industry know, because ChatGPT is what put a model in front of everyone. Reasoning used to be a separate line of models here. It is now a setting on the main one, turned up when a question is worth the extra time and money and left alone when it isn’t.

Claude, from Anthropic. Several sizes ship side by side rather than one flagship, from a small fast model up to the top of the range. Strong on long documents and on code, and common inside large companies.

Gemini, from Google. The strongest of this group on images, audio and video, and wired through Gmail, Docs and the rest of Workspace, which is how most businesses meet it first.

Grok, from Elon Musk’s lab, now part of SpaceX. It sat outside the serious conversation for a while and now sits at or near the top of the general capability rankings, with a business product to match.

The second group publishes the model file itself, so it can be downloaded and run on machines the company controls. DeepSeek and Qwen, from Alibaba, are the two a business is likeliest to meet by name. Kimi, from Moonshot, and GLM, from Z.ai, sit alongside them, and the top of that group has been Chinese for a while now. Mistral, the French lab, publishes files and sells hosted access to both, and turns up wherever rules about which country the data sits in are doing the deciding. Llama, from Meta, is the reason published weights went mainstream at all. It is no longer the top of this group, and Meta has started keeping some of its strongest models closed.

Open-source vs. closed models is where that second column actually gets weighed up: what the licence permits, where the data goes, which cost shape fits.

None of these names are fixed. Labs that were footnotes 2 years ago are in the top 5 now, most of the second group barely registered in a business conversation then, and at least one name above will have moved again within a year. The grouping survives the reordering. The list doesn’t.

Tiers within each family

Each lab ships at roughly 3 sizes. The names change every 6 to 12 months. The pattern doesn’t.

Top tier. The smartest, slowest, most expensive. The models that work through a problem before answering live here. Used for hard reasoning, long-form work, anything where quality matters more than speed or cost per call.

Workhorse tier. Meaningfully cheaper, almost as good for most everyday tasks. This is what most live systems actually run on.

Small and fast tier. Fast, cheap, less capable. Used for high volume and low difficulty: classification, routing, simple extractions, anything that has to run millions of times a month.

Across the big families, the small tier costs somewhere between a tenth and a twenty-fifth as much per page of text as the top tier, and answers in about a second where the top tier can take tens of them. Put a page in and take a short answer out and the top tier costs around a US cent, the small tier a small fraction of one. 100 of those a day and the difference is invisible. 100,000 a month and it is tens of dollars against about $1,000. Which of those two a given workflow lands on is arithmetic that can be done before anything is built.

That is why the tier question is usually settled by the job rather than by preference. Cheap model on the volume, expensive model on the exceptions, is the shape most of this work settles into.

Example

Say every supplier invoice that arrives gets read, matched against a purchase order and flagged where the two disagree: 40,000 a month, the same short job every time. The small tier reads and matches them for a few tens of dollars, fast enough that nothing queues, and wrong rarely enough that the errors land in the flagged pile a person is reading anyway. The top tier would hold the queue up for a difference in accuracy nobody downstream would see; at this volume the money is small either way, the waiting is not. The few hundred flagged invoices, where the disagreement is genuinely ambiguous, are worth the top tier’s time.

The whole curve also shifts down over time. The frontier model of the previous generation is roughly the small, fast tier of the next one, at a fraction of the price. Any cost or capability assumption locked in today goes stale fast.

What “best at” actually means

Marketing pages and benchmark tables say things like “best at coding”, “most creative”, “top of the leaderboard”.

• The numbers change every few months as new models ship. Today’s leader is often next quarter’s middle of the pack.
• Benchmarks are necessarily generic. They measure standardised tasks that may or may not resemble the job the model would actually be put on.
• They don’t account for prompt quality, context, or how the model is wired into the work around it. The same model in a thoughtful workflow beats a smarter model wired badly.
• They don’t capture taste: how a model phrases things, handles ambiguity, hedges or commits, lays out an answer meant to be pasted into a document. For real work, taste is most of the experience.

Example

Say a lab announces its new model is state of the art on a coding benchmark, 2 points ahead of the nearest rival. What that claim licenses: on that benchmark, on that day, run by the lab, the model is at the top. What it does not license: that 2 points is visible on the code any particular team writes, that it survives the next release from a rival, or that the benchmark resembles the team’s work. 2 points on a leaderboard and a decision to change the model wired into everyone’s editor are not facts of the same size.

For most everyday business tasks, drafting, summarising, extracting, classifying, answering questions over documents, the top few models are interchangeable in any way a user would notice. The differences bite at the edges: extremely long documents, hard reasoning, specialised code work, images and audio, less-resourced languages, very strict instruction-following.

Where the differences come from

The labs train on heavily overlapping material and read each other’s research, which is most of why the models land so close together. What separates them is mostly house style, and it comes from two places: what went into the training mix, and what the lab rewarded once the raw model existed. One lab pushes hard on following an instruction to the letter, another on refusing to guess, another on code. Those are choices, made by people, and they are what a team is reacting to when it says a model feels better on this work.

How “which is best” actually settles

Rarely by benchmark. It settles by use: somebody running real work on a model for a few weeks, noticing where it shines and where it frustrates, and forming a taste. Two teams in the same company often land in different places, and both can be right. Done on purpose, on a team’s own past cases rather than over months of live work, that is testing before trusting.

Example

Say the sales team drafts outreach on the model that holds the house tone, and stops arguing about it. Support moves its ticket summaries to a cheaper model because the bill falls by two-thirds and nobody can tell the output apart. Neither team read a leaderboard, and either default may move again when a new model ships.

Which is also why the choice of model is rarely what decides whether an AI project works. Between the top few, the model is the part of the system least likely to be the problem.

Common questions

Is ChatGPT or Claude better?

Neither is better at everything. For everyday drafting, summarising and pulling facts out of documents, the models behind ChatGPT and Claude are close enough that most people would not notice a difference, and the lead between them changes with each release. Claude is known for long documents and code, and beyond that the differences are mostly style, such as how each one phrases things and when it hedges. The reliable way to choose is to run both on your own work for a few weeks and keep the one your team prefers.

Is Gemini better than ChatGPT?

Not across the board. On everyday writing and summarising, the models behind Gemini and ChatGPT are close, and which one tops the rankings changes every few months. Gemini's clearest strengths are images, audio and video, and its place inside Gmail, Google Docs and the rest of Google Workspace. For most teams the choice comes down to which one fits the software they already use and which does better on a trial with their own work.

What does an LLM leaderboard tell you?

An LLM leaderboard ranks AI models against each other, using scores on standard tests or votes from people comparing two answers side by side. It shows which models are strong in general on the day it was measured, and little more: the order changes every few months, the tasks may look nothing like a given team's work, and a lead of a point or two rarely shows up in practice. Use it to pick a shortlist, then test the shortlist on your own past work before choosing.

What is a frontier model?

A frontier model is the newest, largest and most capable model an AI lab offers: the top of its range, and the one its announcements are about. The top models from all the major labs, taken together, are called the frontier. They are also the slowest and most expensive to run, so most everyday business work runs on the cheaper, faster tiers each lab sells alongside them.

Key takeaways

• Every lab has a model at the top of its range: newest, largest, most expensive, the one the announcements are about. That group, across all the labs, is the frontier. Most work never touches it.
• The useful split is how the model reaches you. GPT, Claude, Gemini and Grok are reached through an API or a product. DeepSeek, Qwen, Kimi, GLM, Mistral and Llama publish the file itself.
• The list reorders every year. Labs that were footnotes in the last reshuffle sit in the top five now, and the strongest published files come mostly from Chinese labs rather than from Meta.
• Each family ships at roughly 3 sizes: a top tier, a workhorse that takes most of the everyday load, and a small fast tier for high volume. The names change every six to twelve months. The pattern doesn't.
• The small tier costs somewhere between a tenth and a twenty-fifth as much per page of text as the top tier and answers in a fraction of the time. Cheap model on the volume, expensive model on the exceptions, is the shape most work settles into.
• Capability moves down the cost curve. The frontier model of one generation is roughly the small, fast tier of the next, at a fraction of the price.
• For everyday drafting, summarising, extracting and classifying, the top few models are interchangeable in any way a user would notice. Differences show up at the edges: very long documents, hard reasoning, code, images and audio, less-resourced languages.
• Benchmark claims age in months and rarely match the job in front of a team. Where models genuinely differ is house style, and that shows up in use rather than on a leaderboard.