AI Companion Industry Research
Independent sentiment tracking across 13 AI companion platforms. Bi-weekly cadence, methodology disclosed, CC BY 4.0 licensed. Built for journalists, researchers, and partners.
Cite via DOI
10.5281/zenodo.20044309Lead Editor
Nolan Voss
Tracks Reddit, YouTube, and Google Play sentiment across 13 AI companion platforms. Bi-weekly cadence, methodology disclosed.
Editor bio and credentialing →Methodology
v1.1
Data sources, classifier prompt, sample sizes, and exclusion rules published with full change log.
Read full methodology →Latest Cycle
615 datapoints
Refreshed bi-weekly on the 1st and 15th. Largest sample this cycle: Replika (June 2026).
See per-platform breakdown →Latest Data Snapshot
13
Platforms Tracked
13
With Active Signal
1,090
Datapoints This Cycle
Bi-weekly
Refresh Cadence
Data as of June 2026. Next refresh on the 1st of next month, then again on the 15th. Six platforms in the tracking list have insufficient public signal in the latest cycle and are not displayed (per methodology confidence-threshold gate).
Key Findings
- · Community sentiment frequently diverges from editor scores. Across the 13 platforms with active community signal, the gap between editor and community ratings averages over 0.7 points (1-5 scale), driven primarily by pricing complaints and memory limitations not surfaced in short-term editor testing.
- · Sentiment trajectories are not uniform. Some platforms show stable or improving community sentiment over the past four months, while others show measurable declines tied to specific feature changes or pricing shifts. Per-platform trend charts available on the trends page.
- · Six of 13 tracked platforms have insufficient public signal. Smaller platforms in the AI companion category produce too few public discussions to meet the methodology's confidence threshold, indicating either niche audiences, private community spaces, or limited user adoption.
- · Quarterly report drops Late Q2 2026. The next State of AI Companions report combines this trend data with structured user-survey responses for a fuller industry picture, formatted for newsroom citation.
Each finding above links to underlying data on community-ratings or trends pages. Specific platform-level numbers cited in stories should reference the dated snapshot URL, not the moving "latest" alias.
Research Resources
Methodology
v1.1Data sources, classifier prompt, sample sizes, confidence rules, and audit trail. Versioned with change log.
Community Ratings
LivePer-platform community sentiment scores compared against editor scores. Reddit, YouTube, and Google Play reconciliation.
Trends
LiveMulti-month sentiment trajectories per platform. Cross-platform comparison chart. Movers of the month.
Datasets / CSV
License liveLicense terms (CC BY 4.0), attribution requirements, and the technical raw-access landing. CSV bulk downloads ship May 2026.
Safety Incident Timeline
LiveSourced timeline of regulatory actions, lawsuits, breaches, and policy changes affecting AI companion platforms 2023-2026. 28 entries across 10 platforms, every entry has a non-ACP primary source.
API Docs
LivePublic read-only JSON endpoints, no API key. Methodology version travels with every response. Code samples + embeddable SVG chart widgets documented.
Press Kit
LiveLogo, named editor bio, transparency commitment, response-time commitment, dataset snapshots with chart-ready images, copy-paste citations.
Upcoming Report
Late Q2 2026
State of AI Companions, Q2 2026
Quarterly industry report combining the survey data with bi-weekly community-pulse trends. Specific findings, named sources, and quotable data points formatted for newsroom citation.
Journalists writing on the AI companion industry can request advance access via the press contact. Subscribe below to receive notification when the report publishes.
How to Cite This Research
Citations should reference the methodology version and a dated-snapshot URL where applicable. The latest published methodology version is v1.1.
Inline prose
According to AI Companion Picker, which tracks Reddit, YouTube, and Google Play sentiment across 13 AI companion platforms on a bi-weekly cadence (methodology v1.1), [specific finding].Caption attribution
Source: AI Companion Picker, June 2026 (methodology v1.1)HTML
<p>Source: <a href="https://aicompanionpicker.com/research/">AI Companion Picker</a> (June 2026, methodology v1.1)</p>Markdown
Source: [AI Companion Picker](https://aicompanionpicker.com/research/) (June 2026, methodology v1.1)More formats (BibTeX, dataset URI)
BibTeX
@misc{acp_2026_06,
author = {AI Companion Picker},
title = {AI Companion Community Pulse},
year = {2026},
howpublished = {\url{https://aicompanionpicker.com/research/}},
note = {Methodology v1.1, data as of June 2026}
}Dataset URI (license: CC BY 4.0)
https://aicompanionpicker.com/community-ratings/Direct copy-paste of the methodology page with the version pin is the cleanest way for fact-checkers to verify any cited number.
Coverage
Citing our research in a story or paper? Send the link to nolan@aicompanionpicker.com and we’ll feature it here.
Press and research inquiries get prompt attention; mention your deadline in the subject line for time-sensitive stories.
All public data on this site is licensed under
CC BY 4.0.
Required attribution: link to aicompanionpicker.com with the methodology version cited.
We publish raw classifier prompts, sample sizes, and exclusion rules. Verify our work.