How AI Search Is Recommending Online Therapy: Monthly Trends

Executive Summary

Talkspace remains the category leader in October 2026, holding 75.8% valid recommendation coverage. The gap to the next brand, BetterHelp at 64.2%, stands at an 11.6-point lead. Leadership was not stable this month: Talkspace moved up 8.7 points from September 2026, its sharpest single-month move in the series, extending an upward streak to two months.

Grow Therapy posted the largest baseline-to-current rise in the category, up 12.3 points since July 2026 to 51.7% coverage, completing a three-month upward streak and holding the clear third position. Talkspace's own gain, up 9.1 points since July 2026, was concentrated in placement: the brand recorded 71.2% top-three coverage and 27.3% rank-one coverage in October 2026, both series highs. BetterHelp also rose significantly, up 8.5 points since July 2026 to 64.2% and up 6.5 points from September 2026.

The sharpest decliner on the baseline-to-current view is Online-Therapy.com, down 7.6 points since July 2026 to 28.1% coverage, a three-month downward streak. Regain fell 6.8 points to 11.4% over the same span, also a three-month streak, and 7 Cups fell 6.5 points to 6.6% across a three-month decline.

Against the prior month, three brands moved significantly upward (Talkspace, BetterHelp, Amwell) and Grow Therapy continued rising, while four brands are classified as significant decliners across the full July-to-October series. The category summary for October 2026 is mixed.

Each monthly run begins with 800 prompt-surface observations (441 unique questions in July 2026, 507 in August 2026, 454 in September 2026, and 471 in October 2026) across the benchmark's defined AI/search surface universe. Of those, 800 mentioned a tracked brand or competitor in every measured month. In October 2026, 762 were relevant and 38 were irrelevant, compared with 741 relevant and 59 irrelevant in July 2026. The public metrics use the 545 observations in October 2026 and 510 observations in July 2026 that survive both qualification stages.

Key Findings

AI Response Inconsistency Alerts

Questions This Section Answers

• Which online therapy topics produced the most conflicting AI answers this month?
• How did AI platforms disagree about whether BetterHelp and Talkspace accept insurance?

The benchmark detected 26 critical and high-severity factual inconsistencies across 6 AI platforms in October 2026. Every listed conflict carries high confidence. The conflicts cluster around insurance eligibility, session pricing, and care-model availability, and they matter commercially because a buyer researching online therapy on one surface can receive the opposite answer on another.

BetterHelp

AI platforms provided conflicting information about whether BetterHelp accepts insurance. When asked how much online therapy costs per session, Google AI Mode stated that BetterHelp has a $23 copay with limited plan eligibility, implying it accepts some insurance, citing Blossom Health, Thervo, and HelpGuide.org, while Copilot stated that BetterHelp is always out-of-pocket and does not take insurance, citing TherapyExplained, The Psych Nerd, and SIIT. The flagged sources split on this, with the company's comparison page describing insurance coverage options that vary by state and plan while ChoosingTherapy.com called BetterHelp almost exclusively cash-pay.

Responses also differed on insurance coverage more broadly. When asked whether BetterHelp is the best online therapy, Copilot stated there is no insurance coverage and you pay out of pocket, while Google AI Overviews described limited insurance acceptance depending on the state. Copilot's cited source VirtualTherapyApps stated no insurance is accepted, while BetterHelp's own site described most major insurance providers as in-network with availability varying by state and plan.

A third conflict surfaced on insurance acceptance. When asked what the best psychotherapist platform is, Gemini stated BetterHelp does not accept health insurance directly, while Copilot stated insurance is limited, implying some acceptance. TeleRanked was cited on the Copilot side stating not billed to insurance, and ChoosingTherapy.com on the same side stated the platform takes insurance.

A fourth conflict concerned insurance acceptance again. When asked what the best therapy sites are, Copilot stated insurance was not accepted, while Gemini stated insurance accepted was limited. VirtualTherapyApps and Healthcare Business Today were both cited stating no insurance accepted.

A fifth conflict involved insurance acceptance. When asked what the best online therapy company is, Gemini stated limited direct insurance billing, while Copilot stated the platform does not accept insurance. VirtualTherapyApps was cited on the Copilot side stating no insurance accepted.

A sixth conflict concerned insurance acceptance. When asked what the best platform for therapy is, Copilot stated no insurance is accepted, while Google AI Mode stated insurance compatibility varies. Healthcare Business Today was cited on the Copilot side stating no insurance accepted, and Talkspace.com was cited on the Google AI Mode side describing coverage by most major insurance plans.

A seventh conflict involved insurance coverage. When asked what the best online therapy options are, ChatGPT stated insurance availability varies by plan, state, and therapist, implying it may accept insurance, while Copilot stated insurance coverage was no. TeleRanked was cited on the Copilot side stating not billed to insurance, and Healthline was cited on the ChatGPT side describing a range of insurance coverage.

An eighth conflict concerned insurance acceptance. When asked whether BetterHelp or Talkspace is better, ChatGPT stated BetterHelp accepts insurance through 125+ plans with coverage varying by plan and state, while Google AI Overviews stated BetterHelp does not generally accept health insurance. The company's own cost page was cited describing copays averaging about $23 per session for eligible members, and ChoosingTherapy.com was cited stating BetterHelp is almost exclusively cash-pay.

A ninth conflict involved insurance acceptance. When asked what the best therapy sites are, Google AI Mode stated the platform generally does not bill insurance directly, while Gemini stated insurance accepted was limited. ChoosingTherapy.com was cited stating the platform takes insurance.

A tenth conflict concerned psychiatry availability. When asked whether BetterHelp or Talkspace is better, ChatGPT stated psychiatry is available through UpLift, while Google AI Mode stated BetterHelp focuses exclusively on therapy and does not provide psychiatric care. ChoosingTherapy.com was cited stating BetterHelp focuses only on talk therapy, and BetterHelp's own cost page was cited describing psychiatry services through UpLift.

An eleventh conflict involved insurance coverage. When asked whether BetterHelp is the best online therapy, Copilot stated there is no insurance coverage and you pay out of pocket, while Gemini stated expanding select in-network insurance options. VirtualTherapyApps was cited on the Copilot side stating no insurance accepted.

A twelfth conflict concerned insurance coverage. When asked whether BetterHelp is the best online therapy, Copilot stated no insurance coverage and out-of-pocket payment, while Google AI Mode stated insurance acceptance has rolled out in select states and plans. Healthline was cited stating the platform accepts insurance in some states, and VirtualTherapyApps was cited stating no insurance accepted.

A thirteenth conflict involved insurance acceptance. When asked what the best online therapy that accepts insurance is, ChatGPT stated BetterHelp began accepting insurance through selected in-network providers in 30+ states in 2026, while Google AI Mode stated BetterHelp does not take insurance.

A fourteenth conflict concerned insurance acceptance. When asked what the best online therapy that takes insurance is, Copilot stated BetterHelp has limited but growing insurance coverage, while Google AI Mode stated BetterHelp generally does not accept insurance directly. MindKindly was cited stating coverage is expanding to about a $23 average copay where covered, and HelloBackpack was cited stating the platform generally does not accept insurance.

A fifteenth conflict involved insurance acceptance. When asked what the best online counseling is, ChatGPT stated insurance coverage is more limited, while Copilot stated the platform does not take insurance. MedSwitcher was cited stating BetterHelp does not accept insurance but offers financial aid.

A sixteenth conflict concerned the insurance billing model. When asked what the best online therapy options are, ChatGPT stated insurance availability varies by plan, state, and therapist, while Google AI Mode stated the platform generally operates on an out-of-pocket subscription model rather than billing insurance directly. HelloBackpack was cited stating the platform generally does not accept insurance, and Verywell Mind was cited describing many online therapy companies as in-network with many insurance plans.

Talkspace

AI platforms provided conflicting information about therapist pay rates. When asked which online therapy platform pays the most, Gemini stated $25 to $40 per hour, citing HealthAIDaily, while ChatGPT stated $90 in California, $78 in DC, Massachusetts, Maryland, New York, Virginia, and Washington, and $70 in other states per 60-minute session, citing Talkspace's own therapist recruitment page. Those figures cannot both be correct for a single 60-minute session.

A second conflict concerned therapist pay rate. When asked which online therapy platform pays the most, Gemini stated $25 to $40 per hour, while Copilot stated $30 to $70 per hour.

A third conflict involved insurance acceptance. When asked what online marriage counseling costs, Google AI Mode stated that most major national online subscription platforms do not bill insurance directly, while Gemini stated that unlike many subscription competitors, Talkspace accepts certain major health insurance plans.

A fourth conflict concerned Medicaid acceptance. When asked whether BetterHelp or Talkspace is covered by insurance, Google AI Mode stated Talkspace generally does not accept Medicaid, while Perplexity stated Talkspace accepts some Medicare and Medicaid arrangements depending on the plan and state. CNET was cited stating Medicare and Medicaid do not cover the cost of therapy, and Talkspace's own help center stated it currently does not accept Medicaid or support international health insurance plans.

Brightside Health

AI platforms provided conflicting information about monthly price. When asked what the best online therapy services are, Copilot stated $75 to $95 per month, while Gemini stated $75 or more per week. HealthVetted was cited at $95 per month and TeleRanked at $65 to $100 per week for a different brand on the same page, illustrating how a weekly figure can be misattributed to a monthly line item.

A second conflict concerned cost per period. When asked what the top online therapy companies are, Gemini stated around $75 per week for therapy packages, while Copilot stated $75 or more per month.

A third conflict involved the billing period. When asked what the best online therapist is, Gemini stated around $75 per week for combined care packages, while Copilot stated $75 or more per month.

Grow Therapy

AI platforms provided conflicting information about psychiatry and medication management availability. When asked what the best mental health virtual care is, Google AI Mode stated Grow Therapy connects users with in-network licensed therapists and psychiatrists, while Copilot stated psychiatry and medication management were not offered and the platform is therapy only.

A second conflict concerned session cost with insurance. When asked what the best platform for online therapy is, ChatGPT stated an average of about $21 per session with insurance, while Copilot stated $75 or more per session. Grow Therapy's own how-it-works page was cited stating most clients pay on average $21 per session or less with insurance.

Online-Therapy.com

AI platforms provided conflicting information about monthly price. When asked how much money online therapy costs, Gemini stated roughly $65 to $100 per week, billed monthly and usually totaling $260 to $400 per month, while Copilot stated $64 to $108 per month, or $32 to $54 per session. AllOneTherapy was cited describing flat monthly rates often $200 to $400, and MedSwitcher was cited stating the $64 to $108 per month figure for Online-Therapy.com.

Benchmark Context

Questions This Section Answers

• How many of the collected prompts actually qualified for the online therapy benchmark?
• What changed in the category-wide recommendation rate between July and October 2026?

The report separates the raw collection universe from the qualified analysis set. Brand-level recommendation percentages are calculated within the qualified benchmark set.

Benchmark-Level Metrics

These category-wide figures summarize the qualified benchmark independent of individual brand standings.

August 2026 and September 2026 are not shown as table columns. Recommendation-shaped answer share moved from 48.0% in July 2026 to 43.5% in August 2026 to 43.2% in September 2026 before rising to 58.5% in October 2026; valid recommendation shortlist share held near 66% across July through September 2026 before climbing to 76.9% in October 2026. The category's October lift is a sharp reversal of a three-month plateau, not a gradual climb.

AI Recommendation Trend

Questions This Section Answers

• Who leads online therapy recommendations in October 2026, and by how much?
• Which brands gained or lost the most recommendation coverage since July 2026?

Leadership held, but October's gains were concentrated in the top three rather than spread across the category.

Talkspace's 9.1-point gain, BetterHelp's 8.5-point gain, and Grow Therapy's 12.3-point gain since July 2026 each exceed normal month-to-month variation. The category-level increase in October 2026 is not a uniform shift: it comes from three brands rising while four brands declined significantly over the same period.

What Changed This Month

Questions This Section Answers

• Why did Talkspace's coverage jump 8.7 points in October 2026?
• Which brands declined significantly across the July-to-October period?

Talkspace

Talkspace's valid recommendation coverage rose from 66.7% in July 2026 to 67.1% in September 2026 and then to 75.8% in October 2026, a significant increase of 9.1 points across the period and 8.7 points from September 2026 alone. That prior-to-current move is the sharper story and the brand's largest single-month move in the series.

Placement reinforced coverage. Top-three recommendation rate rose from 56.9% in July 2026 to 71.2% in October 2026, and rank-one recommendation rate rose from 16.9% to 27.3%. Raw mention presence rate moved from 96.1% to 98.0%.

Talkspace is not simply appearing more often, it was already present in nearly every qualified observation in July 2026. The change is in where it sits inside those responses. Rank-one coverage of 27.3% against top-three coverage of 71.2% indicates a meaningful share of responses still place the brand below first.

Highest-priority diagnostic: Which surfaces and prompt types converted rising presence into first-place placements between September and October 2026, and which brands held first place in the responses where Talkspace was recommended but not ranked first?

BetterHelp

BetterHelp's valid recommendation coverage rose from 55.7% in July 2026 to 64.2% in October 2026, a significant increase of 8.5 points, and up 6.5 points from September 2026. This is a two-month upward streak.

The brand's top-three recommendation rate climbed from 46.1% in July 2026 to 58.7% in October 2026, a significant rise of 12.6 points. Rank-one recommendation rate moved from 27.1% to 24.0%, a non-significant decline. Raw mention presence rate moved from 84.7% to 88.4%.

BetterHelp gained ground on the recommendation itself while ceding some ground at the top slot. Coverage and top-three placement are both rising faster than rank-one, which points to the brand winning more shortlist positions without winning proportionally more first positions.

Highest-priority diagnostic: In the responses where BetterHelp is recommended in the top three but not first, which brand holds rank one, and does the answer pattern differ by AI surface?

Grow Therapy

Grow Therapy's valid recommendation coverage rose from 39.4% in July 2026 to 43.5% in August 2026, 48.8% in September 2026, and 51.7% in October 2026, a significant increase of 12.3 points across a three-month upward streak.

Supporting metrics moved in the same direction. Raw mention presence rate rose from 49.6% in July 2026 to 62.7% in October 2026, a significant rise of 13.1 points. Top-three recommendation rate rose from 26.1% to 41.5%, and rank-one recommendation rate rose from 10.0% to 16.3%. The brand's valid recommendation count rose from 201 in July 2026 to 282 in October 2026.

This is the largest baseline-to-current gain in the category, and the only brand rising across three consecutive months. The movement is not confined to presence: rank-one coverage rose by more than half its July 2026 level, meaning the brand is converting more of its added visibility into top placement.

Highest-priority diagnostic: Which prompt types and surfaces are producing the rank-one gains, and are they shared with or distinct from the prompts where Talkspace and BetterHelp place first?

Online-Therapy.com

Online-Therapy.com's valid recommendation coverage fell from 35.7% in July 2026 to 33.1% in August 2026, 28.3% in September 2026, and 28.1% in October 2026, a significant decline of 7.6 points across a three-month downward streak. The move from September to October 2026 was only 0.2 points, so the decline is largely a July-to-September story that held flat in the current month.

Presence and placement both weakened. Raw mention presence rate fell from 41.0% in July 2026 to 31.7% in October 2026, a significant drop of 9.3 points. Top-three recommendation rate fell from 11.2% to 5.9%, a significant drop of 5.3 points. Valid recommendation count fell from 182 in July 2026 to 153 in October 2026.

The brand is losing both visibility and recommendation credit, which distinguishes it from a pure ranking problem. Its net sentiment score held at 0.9 in both months, so the decline is not associated with a shift in how positively the brand is described when it does appear.

Highest-priority diagnostic: Is the presence decline concentrated in specific AI surfaces, and which brand is absorbing the shortlist positions the brand no longer holds?

Regain

Regain's valid recommendation coverage fell from 18.2% in July 2026 to 14.8% in August 2026, 11.6% in September 2026, and 11.4% in October 2026, a significant decline of 6.8 points across a three-month downward streak. The move from September to October 2026 was 0.2 points.

Raw mention presence rate fell from 22.9% in July 2026 to 16.0% in October 2026, a significant drop of 6.9 points. Top-three recommendation rate fell from 6.3% to 4.4%, a non-significant move. Rank-one recommendation rate moved from 2.2% to 2.0%, effectively flat. Valid recommendation count fell from 93 in July 2026 to 62 in October 2026.

Regain's rank-one position resisted the decline even as its broader coverage shrank. The brand lost shortlist slots more than it lost first-place placements, so the drop is concentrated in the middle of the recommendation range.

Highest-priority diagnostic: Which prompts still return Regain in the top three, and are they concentrated on a narrow set of surfaces or evenly distributed?

7 Cups

7 Cups' valid recommendation coverage fell from 13.1% in July 2026 to 10.4% in August 2026, 7.2% in September 2026, and 6.6% in October 2026, a significant decline of 6.5 points across a three-month downward streak.

Raw mention presence rate nearly halved, from 17.2% in July 2026 to 8.6% in October 2026, a significant drop of 8.6 points. Top-three recommendation rate rose from 2.2% to 3.7%, a non-significant move, and rank-one recommendation rate held near flat at 1.4% to 1.5%. Valid recommendation count fell from 67 in July 2026 to 36 in October 2026.

The brand is being surfaced in fewer responses at the same time that its share of top-three placements edged up. Those two facts point in opposite directions and need prompt-level inspection before either is read as a trend. The valid recommendation count of 36 in October 2026 is small enough that the percentage carries more uncertainty than for larger brands in this benchmark.

Highest-priority diagnostic: Which prompts stopped returning 7 Cups entirely, and were those discovery-stage prompts where the brand previously appeared without a top placement?

Cerebral

Cerebral's valid recommendation coverage fell from 7.6% in July 2026 to 7.0% in August 2026, 4.6% in September 2026, and 2.9% in October 2026, a significant decline of 4.7 points across a three-month downward streak. Raw mention presence rate fell from 13.5% to 7.2% over the same period, a significant drop of 6.3 points, and top-three recommendation rate fell from 1.4% to 0.2%, also a significant decline. Valid recommendation count fell from 39 to 16, and net sentiment softened from 0.6 to 0.4.

Talkspace records 75.8% valid recommendation coverage in October 2026, while Cerebral stands at 2.9%. Talkspace also converts presence into placement at 27.3% rank-one coverage, while Cerebral's rank-one rate has remained at 0.0% in every one of the four measured months. The more important issue is what that gap means for Cerebral's recommendation position: the brand is not merely behind the category leader on volume, it has never been placed first in a qualifying AI response across the full July-to-October series, even as its broader coverage continues to shrink. With only 16 valid recommendations remaining in October 2026, the brand's position is both the weakest on record and the most fragile to measure, and the underlying cause of the zero rank-one rate remains unexamined.

Highest-priority diagnostic: Which prompts still return Cerebral at all, and what attributes or evidence sources appear in the responses where it is mentioned but not recommended — and why has rank-one placement remained at zero across all four measured months while Talkspace's rank-one rate reached 27.3%?

Amwell

Amwell's valid recommendation coverage moved from 10.2% in July 2026 to 10.2% in August 2026, 6.8% in September 2026, and 12.3% in October 2026, ending up 2.1 points on the baseline view but up 5.5 points from September 2026, a significant prior-to-current move.

Top-three recommendation rate rose from 2.8% in July 2026 to 6.1% in October 2026, a significant rise of 3.3 points. Raw mention presence rate held essentially flat at 17.2% to 17.6%. Rank-one recommendation rate moved from 0.8% to 0.5%. Valid recommendation count rose from 52 in July 2026 to 67 in October 2026.

Amwell's September 2026 dip recovered fully in October 2026, and the recovery came with a broader top-three footprint. The brand's presence has been stable while its recommendation credit has swung, which makes it the clearest example in this series of presence and recommendation behaving as separate signals.

Highest-priority diagnostic: Which surfaces and prompts account for the top-three gains in October 2026, and do they overlap with the surfaces where the brand is mentioned but not recommended?

Brightside Health

Brightside Health's valid recommendation coverage held at 41.4% in July 2026 and 41.3% in October 2026, effectively flat across the period. Intermediate readings were 38.1% in August 2026 and 41.8% in September 2026.

Raw mention presence rate moved from 47.1% in July 2026 to 44.6% in October 2026, a non-significant decline. Top-three recommendation rate moved from 14.9% to 15.4%, essentially flat. Rank-one recommendation rate moved from 0.8% to 0.4%. Valid recommendation count rose from 211 in July 2026 to 225 in October 2026.

Brightside Health is the fourth-ranked brand by coverage and the steadiest of the mid-tier. Its rank-one rate remains low relative to its coverage, which means the brand's visibility is concentrated in shortlist positions below first place.

Highest-priority diagnostic: Given coverage and top-three are stable but rank-one is minimal, which prompts place the brand first, and how do those prompts differ from the shortlist prompts?

Calmerry

Calmerry's valid recommendation coverage rose from 4.7% in July 2026 to 5.9% in October 2026, an increase of 1.2 points that is not classified as significant. Top-three recommendation rate rose from 1.2% to 3.9%, a significant move on the baseline view. Raw mention presence rate moved from 9.0% to 7.7%. Valid recommendation count rose from 24 in July 2026 to 32 in October 2026.

Calmerry is the smallest-coverage brand in the tracked set apart from Cerebral, so its top-three movement should be read with the count in view: the significant top-three rise reflects 21 placements in October 2026 against 6 in July 2026.

Highest-priority diagnostic: Which individual prompts are producing the top-three placements, and are they stable enough to hold month over month?

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent clusters did the qualified observations fall into?
• What can the benchmark measure about how AI frames online therapy pricing or brand comparisons?

All 545 qualified observations in October 2026 fell into the Brand Recommendation cluster, mirroring July 2026 when all 510 qualified observations also fell into the same cluster. There were zero qualifying observations in the Pricing & Value or Multi-Brand Comparison clusters in either month.

The practical consequence is that the public benchmark can describe who AI recommends in this vertical but not how AI frames cost, value, or head-to-head comparison. The inconsistency alerts above show that pricing and comparison claims are being made by AI platforms, and are conflicting across them, but those answers did not meet the qualification bar for inclusion in the brand-level coverage metrics in the measured periods.

Brand Opportunity Summary

Questions This Section Answers

• Which online therapy brands are rising and which are declining in October 2026?
• What is the highest-priority diagnostic question for each brand's recommendation position?

The benchmark identifies where attention is warranted; a company-level analysis is needed to explain why.

Evidence Behind the Benchmark

The aggregate metrics are built from prompt-level observations (query, surface, recommendation outcome, rank, sentiment, and citations where exposed). Company-level analysis can go deeper into prompt, competitor, surface, and evidence patterns. Source presence is not automatically treated as proof of causation.

About This Benchmark

This report is part of the LLM Authority Index AI Visibility Market Discovery research program.

• AI Visibility Industry Market Methodology
• AI Visibility Industry Market Metrics
• AI Visibility Industry Market Standards

Report-Specific Interpretation Notes

• Small observation counts for brands like Cerebral (16 valid recommendations) and 7 Cups (36 valid recommendations) mean their percentages carry more uncertainty than larger brands in this benchmark.
• The qualified denominator (545 observations in October 2026) is smaller than the raw collection (800 prompts); percentages reflect the qualified set only.
• Movement between months is directional analysis, not proof of cause. A rising or falling coverage rate identifies where change is happening, not why.
• Presence, coverage, top-three placement, and rank-one placement are distinct signals. A brand can hold high presence while converting only part of that presence into top-ranked placement, as this month's Talkspace and Amwell patterns both illustrate.

Next Step

The Public Benchmark Shows Where a Brand Is Winning or Losing. A Company-Level Audit Shows Why.

The aggregate percentages in this report raise questions that the public benchmark cannot answer: which high-intent prompts is a brand winning or losing, which competitor takes the recommendation when a brand loses, what attributes do AI responses associate with each option, and which external sources shape those answers. For Talkspace's late-series gains, Grow Therapy's three-month climb, or Cerebral's persistent absence from rank-one placements against a 75.8% leader, the underlying prompts, surfaces, and evidence sources remain unexamined.

The public percentage cannot identify the prompts, competitors, or sources producing the result. A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy.

Request an AI visibility audit