Written by us, so read it with that in mind. Every claim below is sourced, dated July 2026, and correctable. If we got something wrong, email us and we will fix it.
Simi is a real product by a serious team. They generate whiteboard explainers in under 20 seconds, they bet on pedagogical pacing when everyone else was chasing cinematic AI video, and they validated this category loudly enough that we are all here. If raw one-shot generation speed is what you care about most, Simi is very good at it.
We built Sketchie because of what happens after the first generation.
Simi’s own launch language says it: their output is "a finished video, not a draft." Sketchie’s entire premise is that no first generation is finished, so every video ships as editable scenes.
Our column is what we claim about our own product. Their column is what their published materials say as of July 2026, or "not documented" where they are silent. We do not guess at their roadmap or their limits.
| Sketchie (our claims) | Simi (published, July 2026) | |
|---|---|---|
| Output | video made of editable scenes | "a finished video, not a draft" (their launch copy) |
| Fixing one scene | edit and re-render that scene, seconds, cents | no partial regeneration documented |
| Length control | 30-second steps, plus auto length | preset lengths, 1-minute steps |
| Max length | 6 minutes, chaptered series beyond that (by design) | up to 1 hour in one shot |
| Voices | voice library with audience-designed voices, plus bring your own ElevenLabs or Fish Audio voice ID | male or female |
| Languages | 23 languages on our own voice stack, long-tail coverage behind the same picker | 80+ languages |
| Pricing | $19.99 to $499.99/mo, same ladder, plus a full-featured 7-day trial | comparable tiers, $12/mo entry, up to $499.99/mo |
| Custom branding | included in Business | $499.99/mo tier |
| API access | included at 1x on Max and Business | 3x credit markup |
| Generation speed | about a minute | under 20 seconds |
| Scene editor | yes, the core of the product | none visible in the app as of their launch |
| Ending | retrieval question, pause, recap (research-backed) | no retrieval ending documented |
A fair caveat, stated plainly: they can ship anything in that column whenever they choose, and with $3.5M they probably will. Our bet is not that they stay still. Our bet is that editability works best when the video is built as an editable document from the first render, which is how Sketchie is architected, not a feature we added.
Simi sells one-shot videos up to an hour long. We cap at 6 minutes and turn longer material into a chaptered series. That is not a limitation we are spinning. Analysis of 6.9 million edX learning sessions found engagement is near total under 6 minutes, drops to about half at 9 to 12 minutes, and collapses past 12. A 1-hour one-shot explainer is a product decision made against the strongest engagement dataset that exists for this format. We made the opposite decision, on purpose.
Simi is faster at first generation. Under 20 seconds is genuinely impressive and we are not going to pretend otherwise. Our position: the metric that matters is time to a video you will actually publish. That includes every revision. One wrong sentence on a one-shot tool means another full generation and another full spend, as many times as it takes. On Sketchie the second draft takes seconds and costs cents. Try the same prompt on both and time the whole loop, not just the first render.
Honesty section. As of July 2026:
We generated the same prompts on both products and put the results side by side. No cherry-picking: the board includes the ones where Simi looks good.