RUNNING GUIDES · AI COACHING

ChatGPT as a running coach: where it works, where it breaks

Plenty of runners are training off a chatbot plan right now, and some are doing fine. This is the honest map of when that goes well, when it quietly goes wrong, and what the peer-reviewed evidence actually says. It applies to Claude and Gemini just as much, including the models that power our own app.

UPDATED JUL 18, 2026 · 8 MIN READ

Where ChatGPT is genuinely good

Let's start with the credit, because it's real. As a running encyclopedia, ChatGPT is better than most of what you'll find by searching: it explains lactate threshold, supercompensation and 80/20 intensity distribution clearly, patiently, and at whatever depth you ask for.

Decoding jargon. “What does 6×800m at 5K pace actually mean?” Perfect use.
One-off questions. Fueling for a 90-minute long run, what to do the week after a half marathon, how to come back from a cold. Good, mainstream answers.
Reformatting. Turning a plan you already trust into a weekly calendar, or adapting session wording for a treadmill. Mechanical, and it nails it.
A sounding board. Thinking out loud about goals and trade-offs, as long as you remember it will mostly agree with you (more on that below).

What the research found

This isn't just vibes. A peer-reviewed study in the Journal of Sports Science and Medicine (2024) had coaching experts formally evaluate ChatGPT-generated training plans for recreational runners. The headline finding is in the paper's own title: the plans were not rated optimal. Quality improved when prompts included more information about the runner, but even well-fed prompts didn't reach the standard the experts wanted.

Note what that study measured: the plan on day zero, the thing chatbots are best at. Coaching mostly isn't writing the plan. It's the six weeks that follow, and that's where the structural problems live.

The four ways it breaks as a coach

01It can't see you. ChatGPT has no feed of your runs, sleep, HRV or resting heart rate. It doesn't know your last three easy runs drifted 20 seconds per km slower at the same heart rate, the classic early-overreaching signature. You'd have to notice that yourself and paste it in; the runners who most need a coach are precisely the ones who don't notice.
02It doesn't reliably remember. Chat memory features are shallow and easy to silently lose. The calf niggle you mentioned three weeks ago is not in the room when this week's plan gets written. That's the exact opposite of what makes a coach worth having.
03It agrees with you. Propose adding a second interval session to an already-hard week and you'll usually get an enthusiastic yes with a plan attached. Chatbots are trained to be helpful, and in training, unconditional helpfulness is a hazard: the most valuable coaching sentence is no, not today, said unprompted.
04It states numbers with unearned confidence. Paces, mileage jumps and race predictions come out fluent and specific whether they're grounded or not. A 10% weekly mileage rule applied to your bad ankle history is not a pace table problem. It's a judgment problem, and fluency hides the difference.

None of this is a dig at any one model. Claude and Gemini share every one of these limits, and Rayvik itself is built on this class of models. The point is architectural: a chatbot is a brilliant reasoning engine with no eyes, no memory and no brakes. Coaching needs all three bolted on.

The three missing pieces. None of them are about intelligence: they're about what the chatbot is connected to, what it retains, and what it's willing to refuse.

FROM THE FOUNDER

This page is personal. Before my first marathon I gave ChatGPT my half-marathon PB of 1:47 and asked it to plan my race. It ran the conversion formulas and told me, with total confidence, that sub-4:30 was no problem. It saw one number and drew a straight line: it had no idea my mileage was half-marathon mileage, that the strength work wasn't there, or that race day would be hot. I cramped at 26 km and walked in at 5:30, an hour past my floor. Nobody who has actually blown up at 26 km would have given that answer with that much confidence.

The memory problem isn't abstract either. When I kept using chatbots for training afterwards, I'd mention a tight calf, and two weeks later they'd have me doing intervals on it. I was the one carrying all the context. I was the memory.

— Kai, founder of Rayvik · the full story

If you use it anyway: do it like this

Chatbot coaching costs nothing to try (the health-data connectors are another story, and paid), and used carefully it beats no structure at all:

Front-load your context every time: age, weekly mileage, recent race times, injury history, available days. The JSSM study found more input measurably improves plan quality.
Ask it to argue against you. “What's the strongest case that this plan is too aggressive for me?” cuts through the agreeableness better than asking for approval.
Keep a canonical training log elsewhere and paste the last two weeks in before any plan revision. You are the data feed now, so be a good one.
Never take a pace target from a chatbot into a hard session without sanity-checking it against a recent race result. If they disagree, trust the race.

If that list feels like a part-time job, that's the honest summary. You can be the eyes, memory and brakes for a free chatbot, or you can use a system where those are built in. That's the actual dividing line in the AI running coach category, and it's worth understanding before you pay for anything (or before you shop the app alternatives).

ChatGPT running coach FAQs

Can you use ChatGPT as a running coach?

You can, and for one-off questions it's genuinely useful. As an ongoing coach it has structural problems: it can't see your training data, doesn't reliably remember your history, and tends to agree with whatever you propose. Coaching experts who formally evaluated ChatGPT-generated training plans rated them below optimal standards.

Does ChatGPT give good running advice?

On general principles (what threshold pace is, why easy runs matter, how to structure a taper), yes, usually. The risk isn't the concepts, it's the specifics: paces and weekly structures are generated confidently whether or not they fit you, and the plan doesn't change when your recovery does.

Which AI is best as a running coach?

Among chatbots, the differences are smaller than people hope. Claude, ChatGPT and Gemini all share the same structural limits: no data feed, weak memory, agreeable by default. The meaningful choice isn't between chatbots; it's between a chatbot and a purpose-built coaching app that reads your actual training and recovery data.

Is Claude better than ChatGPT for running plans?

There are real differences. Claude's longer context window tends to keep a multi-week plan coherent for longer before it drifts. But both fail the same way as coaches: they can't see your runs, your sleep or your heart-rate data unless you paste it in every time. Choose either for learning; choose neither as your only source of pacing decisions.

Is there research on ChatGPT training plans?

Yes. A 2024 peer-reviewed study in the Journal of Sports Science and Medicine had coaching experts evaluate ChatGPT-generated plans for recreational runners. The plans were not rated optimal, though ratings improved when the prompts included more information about the runner. That's evidence for both the ceiling and the 'better prompts help' folk wisdom.

What is an AI running coach? →The category map: template apps, chatbots, and adaptive coaches.Runna alternatives, compared honestly →The paid apps people graduate to from chatbot plans.Whoop alternatives for Apple Watch owners →The recovery data a chatbot can't see. You may already collect it.