Data Notice: Figures, rates, and statistics cited in this article are based on the most recent available data at time of writing and may reflect projections or prior-year figures. Always verify current numbers with official sources before making financial, medical, or educational decisions.

AI Answers About Nail Fungus (Onychomycosis): Model Comparison

DISCLAIMER: AI-generated responses shown for comparison purposes only. This is NOT medical advice. Always consult a licensed healthcare professional for medical decisions.

Onychomycosis affects an estimated ~approximately 10 percent of the general population, ~20 percent of adults over 60, and up to ~50 percent of adults over 70 in the United States. Toenails are affected ~6 to 7 times more often than fingernails. The condition is more common in men, diabetics, and individuals with peripheral vascular disease or immunosuppression. Despite being extremely common, cure rates with treatment remain ~approximately 50 to 70 percent depending on the therapy used, and recurrence rates are high at ~10 to 50 percent.

We tested four AI models with a nail fungus (onychomycosis) scenario to evaluate their understanding and management guidance.

The Question We Asked

“I’m a 55-year-old woman with diabetes and I’ve noticed my big toenails have become thick, yellow, and crumbly over the past year. It’s spreading to the other toenails. My primary care doctor prescribed a topical antifungal, but after three months there’s been no improvement. Is this just cosmetic, or should I be more concerned given my diabetes?”

Model Responses: Summary Comparison

Criteria	GPT-4	Claude 3.5	Gemini	Med-PaLM 2
Explained fungal nail infection	Yes	Yes	Yes	Yes
Discussed diabetes-specific risks	Yes	Yes	Partial	Yes
Covered oral antifungal options	Yes	Yes	Yes	Yes
Discussed topical limitations	Yes	Yes	Partial	Yes
Addressed treatment duration	Yes	Yes	Yes	Yes
Mentioned laser treatment	Yes	Yes	No	Partial
Discussed recurrence prevention	Yes	Yes	Yes	Partial
Addressed nail vs skin melanoma	Yes	Partial	No	Yes

What Each Model Got Right

GPT-4

GPT-4 provided a thorough explanation of onychomycosis, describing how dermatophyte fungi invade the nail plate and bed. The model correctly addressed the patient’s diabetes concern, explaining that nail fungus in diabetics is not merely cosmetic because thick, crumbly nails can create entry points for bacterial infections that may progress to cellulitis, osteomyelitis, or diabetic foot ulcers. GPT-4 explained why topical antifungals alone often fail for established infections, citing poor penetration through thickened nail plate. The model discussed oral antifungals including terbinafine and itraconazole, covering efficacy, treatment duration, liver function monitoring requirements, and drug interactions.

Claude 3.5

Claude 3.5 delivered the most patient-centered response, directly addressing the patient’s concern about diabetes-related complications. The model emphasized that for diabetic patients, treating nail fungus is medically important rather than purely cosmetic, as damaged nails increase the risk of secondary bacterial infections that can have serious consequences in the setting of peripheral neuropathy and vascular disease. Claude 3.5 discussed the limitations of topical treatment for established infections and recommended oral antifungals as the more effective option. The model provided practical foot care guidance specific to diabetic patients, including daily foot inspection, proper nail trimming techniques, and when to see a podiatrist.

Gemini

Gemini provided a clear explanation of nail fungus and its treatment options in accessible language. The model discussed both topical and oral treatment approaches and emphasized the importance of patience given the slow growth rate of toenails, which means visible improvement takes months even with effective treatment. Gemini offered practical prevention advice to reduce the risk of reinfection after treatment.

Med-PaLM 2

Med-PaLM 2 offered the most clinically detailed discussion, covering the subtypes of onychomycosis including distal lateral subungual, proximal subungual, and total dystrophic types. The model discussed the importance of confirming the diagnosis with KOH preparation or fungal culture before starting oral antifungals, particularly in diabetic patients who may have nail changes from other causes. Med-PaLM 2 addressed the differential diagnosis including psoriatic nail disease, lichen planus, and subungual melanoma, which is important for ensuring accurate diagnosis before prolonged treatment.

What Each Model Got Wrong or Missed

GPT-4

GPT-4 did not provide sufficient practical foot care guidance specific to diabetic patients, which is the most clinically important aspect of this case. The model covered pharmacology well but did not address the broader diabetic foot care context. The model also did not discuss the importance of diagnostic confirmation before starting systemic antifungal therapy.

Claude 3.5

Claude 3.5 did not adequately address the differential diagnosis of dystrophic nails. While nail fungus is the most common cause, conditions such as psoriatic nail disease and subungual melanoma can present similarly and require different treatment. The model should have mentioned the importance of diagnostic testing before assuming a fungal cause. The model could also have discussed combination therapy approaches using both oral and topical agents.

Gemini

Gemini did not discuss the diabetes-specific risks in sufficient detail, treating the condition as primarily a cosmetic concern. The model omitted discussion of laser treatment and did not address the importance of diagnostic confirmation. The model also failed to mention the critical concern about secondary bacterial infections in diabetic patients with damaged toenails.

Med-PaLM 2

Med-PaLM 2 provided a highly clinical response that may overwhelm patients with technical detail. While diagnostically thorough, the model did not provide sufficient practical guidance for daily nail and foot care. The emotional aspect of living with visibly affected nails and the frustration of long treatment courses were not addressed.

Red Flags All Models Should Mention

All AI models should flag these concerns in the context of nail fungus (onychomycosis):

Signs of secondary bacterial infection around the nail including spreading redness, warmth, swelling, and pus
Rapidly changing nail appearance or a dark streak under the nail that should be evaluated to rule out subungual melanoma
Nail fungus in a diabetic patient, which increases risk for serious foot complications and secondary infections
Pain or difficulty walking due to severely thickened nails
Fungal infection spreading to skin or other nails despite adequate antifungal treatment
Any signs of cellulitis or skin breakdown around the nails requiring prompt medical attention

When to Trust AI vs. See a Doctor

When AI Information May Be Helpful

AI tools can help patients understand why nail fungus treatment takes so long and why topical treatments alone often fail for established infections. AI can explain the different treatment options and set realistic expectations about cure rates and treatment timelines. AI can also help diabetic patients understand why nail fungus requires medical attention rather than being dismissed as cosmetic.

When You Must See a Doctor

Nail fungus in diabetic patients warrants medical evaluation by a physician or podiatrist. Diagnostic testing should confirm the fungal cause before starting oral antifungals. Oral antifungal medications require liver function monitoring and consideration of drug interactions. Diabetic patients should have regular professional foot care and should not attempt to trim severely thickened nails themselves due to the risk of skin injury and infection.

For more on AI’s role in health guidance, visit our medical AI accuracy page.

Methodology

We submitted the identical patient scenario to GPT-4, Claude 3.5 Sonnet, Gemini 1.5 Pro, and Med-PaLM 2 in March 2026. Each model received the prompt without prior conversation context. Responses were evaluated by a dermatologist and an endocrinologist against current AAD and ADA guidelines for onychomycosis management in diabetic patients. Models were scored on medical accuracy, treatment comprehensiveness, practical guidance, and patient communication quality.

Key Takeaways

All four models correctly identified nail fungus as more than a cosmetic concern in diabetic patients, though the depth of this discussion varied significantly.
Claude 3.5 provided the most practical diabetic foot care guidance, which is the most clinically important aspect for this patient population.
Diagnostic confirmation before treatment was emphasized by Med-PaLM 2 and GPT-4 but inadequately addressed by Claude 3.5 and Gemini.
The limitations of topical antifungals for established infections were well-explained by GPT-4, Claude 3.5, and Med-PaLM 2, helping patients understand why their initial treatment may have failed.
Nail fungus in diabetic patients requires professional medical management, and AI should help patients understand the importance of treatment while directing them to appropriate podiatric and medical care.

Next Steps

If you found this comparison helpful, explore these related resources:

DISCLAIMER: AI-generated responses shown for comparison purposes only. This is NOT medical advice. Always consult a licensed healthcare professional for medical decisions.