Kolbo.AI brings creative models into one production workspace

Kolbo.AI is aimed at creators who already use several generative tools and keep losing time moving between them. Its pitch is a shared workspace for images, video, audio and other media, rather than a new foundation model. The product tries to keep assets and creative context together while users switch between different generators.

Keeping a production together

The public website presents a creative assistant, a director workflow, image and video tools, audio generation, media organization and review features. It also advertises tools for maintaining character and brand consistency. The important question for a working creator is whether these pieces reduce the handoffs between stages of a real project.

One subscription is not the same thing as unlimited production. Model access, credits, generation quality and export requirements still affect the cost of a finished asset. A sensible trial would take one repeatable job, such as a short branded video, through the whole workflow rather than judge the service from a single impressive example.

How the founder changed the development process

Zohar Vanunu says he had no coding skills and had previously worked with developers. His published account describes rebuilding much of the platform with Lovable, which he credits with about 80% of the product, before moving further with Cursor and Claude Code. This is not evidence that no human developer ever touched Kolbo.

The story is useful because it describes an operational change: the founder could ship more changes himself. The percentage is his estimate, not an independent examination of the repository. The project should be understood as AI-assisted development evolving over time.

A $90,035 revenue snapshot

On October 30, 2026, the public TrustMRR profile showed $90,035 in revenue over the preceding 30 days, $39,299 MRR and $363,419 in lifetime revenue. The profile identifies a Stripe API connection as its verification method.

These figures measure different things. The 30-day revenue includes receipts in that window, while MRR normalizes recurring subscriptions. Neither reveals profit after model bills and other operating costs. The screenshots in this listing are CoderHype captures of the public website, including its demonstrations; they are not outputs generated during a paid product test.

Sources and verification

• Official product page
• Development history and original-source references
• Revenue snapshot on TrustMRR

Financial figures are dated observations and may change. Revenue is not profit. Product screenshots were captured from public pages; developer-provided demonstrations visible within those pages remain attributed to the product.