日本語
#1 Joi7.3/10 · from Not publishedTry free
HomeReviewsJoi Review

Joi Review

7.3/10Ranked #20 of 31Operator-reported

NSFW character chat with a large public character catalogue and image generation, aimed at browsing rather than building.

Verdict

Score 7.3/10

NSFW character chat with a large public character catalogue and image generation, aimed at browsing rather than building.

Joi homepage, as it appears to a new visitor
Joi as it looks on desktop. Screenshot supplied by the operator.
  • セッションをまたいで続く長期記憶
  • 相手の画像をいつでも生成
  • 自分だけの相手を一から作成
  • すぐ始められる既製シナリオ
  • 画像は可、動画はまだ
Visit JoiVisit

Scoring breakdown

CategoryWeightScoreRatingWhat it covers
Conversation25%8.7Model quality, roleplay coherence, response latency.
Memory20%8.3Long-term recall architecture and user control over it.
Media20%7.0Image and video generation quality, cost and throughput.
Voice10%4.8TTS quality, live calls, custom voice design.
Value15%6.6Cost per unit of real usage, not headline price.
Trust10%6.4Corporate transparency, refund terms, billing conduct.
Overall100%7.3Weighted average of all categories.

Pricing

Cheapest planNot published Not published on the site
Rolling monthlyNot published
Free tier✘ No
Plans availablePaid
Extra currencyUSD (charged separately for media)

Note that Joi charges for image and video generation using a separate currency it calls USD, on top of the subscription. Your real monthly spend will be higher than the headline price if you generate media regularly.

Features & specifications

Explicit content✔ Yes
Image generation✔ Yes
Video generation✘ No
Voice messages✘ No
Live voice calls✘ No
Group chat✘ No
Create your own character✔ Yes
User personas✔ Yes
Preset scenarios✔ Yes
Memory systemLong-term
Character libraryNot published
Operating companyJoi
JurisdictionNot published

Ready to try Joi?

Rated 7.3/10 · From Not published

Visit JoiVisit

About the author

AS

Andy Smith

Principal Research Analyst · CompanionRank

Andy Smith is Principal Research Analyst at CompanionRank. He has spent the last two years taking apart AI companion platforms, reading their shipped client code, their public API responses and the company filings behind them, because almost nothing in this category can be taken at face value.

More about our team