Genspark on X

The work people describe is concrete: competitor sheets, reports, slides, meeting minutes, image sets, fact-checking and a second brain for notes. A founder has handed accounts out to employees to raise AI awareness; one user's agent on iMessage booked a Las Vegas dinner and handed off to a GenTeam assistant to send the invites — "a team of avengers in my pocket". One person rates document writing precise and error-free, calling the subscription-only access its one drawback.

Praise and doubt are level. Fans are short bursts of thanks and excitement, often about a single feature. The mixed group holds something back: Genspark is slow, a local-government trial counted hours saved but also more time spent checking and fixing, and one former regular now lists it among tools they used to use a lot. The disappointed call the free standard mode unusable, say search listed one company twice, and argue Claude Code has taken the attention.

The reliability posts are the specific ones: a day-and-a-half outage that made a daily user realise how much they relied on it, an illustration that disappeared into a 404 screen, someone weighing whether to file a CVE for the AI browser, and a user who "can't forgive" a checkbox bug that cost six hours.

On credits, people say they drain fast and buying more is expensive, "so scammy"; one cancelled Plus after two months because image generation hit limits, and another warns video burns credits. The wrong-result complaints are one-off comparisons: a 3D model that came out of the wrong object, a search listing two of five companies as the same firm. One paid user says it doesn't read what they mean.

A four-part write-up of GenTeam setup comes from 1 people: agents idle until nudged, a handoff that stalled, permissions unclear.

The kit talk is specific about the division of labour. One widely shared post assigns research to NotebookLM, writing to Claude and "まとまったやつ" — the roundup, or the thing that puts it together — to Genspark. Another user keeps ChatGPT, Flow, Grok and Genspark in the same rotation; a third says ChatGPT, Gemini and Genspark are "enough". One person runs Genspark from Grokbot through its CLI and API key to save credits. One recommends pairing Claude with Genspark because Claude is weak at images and video.

The plain recommendations are short replies naming Genspark or telling someone to use it, including for free translated viewing. The pack comparisons include the doubters: one says Claude Code has eclipsed it and "everyone has forgotten it"; another can't follow everything new. Single pairings name Grok and its bot UX — one fan's Genspark use is slipping because of it — and Codex, where Genspark is recommended as faster for article images.

The verdict requests are short and open: "gensparkってどうなん？" ("how's Genspark?") from someone deciding whether to try it, and a reply asking what someone makes of it after testing the three main AIs. Nobody in these posts answers with a comparison of their own, so the questions sit unanswered on the page.

The existence question is the sharpest: one post asks @genspark_ai directly whether Genspark is "still a thing" beside HybriqsOS and Grok Bot, and another says the crowd that once posted about Genspark and Manus went quiet once Claude Code arrived. The Claude question is a single reply from someone who has barely used Genspark and wants to know whether it is better than what they have.

The wish is tied to a small, concrete habit: they write feedback because Genspark gives credits for it. From that they say the product is still developing and they want it to grow — "まだまだ発展途上なAIだし成長してほしい" ("it's still an AI in development, and I want it to grow").

That is the whole of the wish talk here, so it is told as one person's view rather than as what users want. Nothing in the posts asks for a specific feature, a price change or a fix; there is no shared wishlist to report yet.