Simon Says vs Wideframe: Transcript Assembly vs Semantic Assembly for Video Editors

Two Different Assembly Models

Simon Says and Wideframe both promise to take you from raw footage to a rough cut faster than manual editing. They use fundamentally different methods to do it, and once you understand the difference, the comparison becomes much more straightforward.

Simon Says is built around a transcript-first assembly model. The tool transcribes your footage with high accuracy, you read or search the transcript, you select the quotes and segments you want, and Simon Says assembles those selections into a timeline that exports to your NLE. The transcript is the editing surface. The cut follows the words.

Wideframe is built around a semantic assembly model. The tool indexes your footage by what is on screen and what is being said -- visual content, actions, compositions, objects, emotional content, plus dialogue. You search across all those signals, select the clips you want, and Wideframe assembles a rough cut that exports as a native Premiere Pro project. The visual layer is a first-class citizen of the search and the assembly.

Neither approach is wrong. They serve editors with different problems. The right framing is to figure out which problem you actually have, then pick the tool whose architecture matches.

How Simon Says Works

Simon Says built its reputation on transcription quality and language coverage. The platform supports transcription in many languages, with strong accuracy on professional audio and a polished editing surface for working with the resulting transcripts.

The Simon Says Assembly workflow starts with a transcribed file. You read the transcript, highlight the parts you want to keep, mark them as selects, and Simon Says builds an EDL or XML that you import into Premiere, Final Cut Pro, or DaVinci Resolve. The output is a sequence of cuts that correspond to your transcript selections. Where you cut the text, the timeline cuts. The handoff to NLE is via interchange formats rather than native project files.

This is genuinely useful for content where the story lives in the words. Long-form interviews, documentary subject sessions, podcast videos, oral histories, depositions, and any project where the editorial decisions are mostly "keep this quote, cut that one" map cleanly onto Simon Says's model. The tool removes the manual labor of scrubbing through long recordings to mark in and out points -- you scan the transcript instead.

The limitation is structural. Simon Says assembles from the transcript, which means everything that is not in the transcript is invisible to the assembly. B-roll, reaction shots, silent moments, action sequences, establishing shots -- these have no transcript, so they have no surface to select from. You can layer them in manually after the fact in your NLE, but the AI assembly does not help with that part of the work.

How Wideframe Works

Wideframe takes a broader starting point. Footage is indexed by both visual content and dialogue. Vision models capture what is on screen -- actions, compositions, objects, emotional content, camera motion. Audio models capture transcript and speaker information. Both layers are searchable from the same query interface.

The assembly workflow starts with search. You ask for the kind of content you need in plain language. "Wide shot of the warehouse with no people." "Reaction of someone laughing during the interview." "Handheld walking sequence through the city." "Moment when the guest said the budget was approved." Wideframe ranks results by how well each clip matches the query, drawing on both the visual and the audio layer.

Once you select the clips you want, Wideframe assembles them into a rough sequence. The output is a native .prproj file for Premiere Pro -- bins, sequences, clip metadata, and references all preserved. Open the project in Premiere and the rough cut is there, ready to finish. Footage stays local on your Mac throughout, indexed in place rather than uploaded to the cloud.

The advantage of this model for visual content is direct. B-roll, reaction shots, silent moments, and action sequences are first-class search targets. They have no transcript, but they have rich visual content, and Wideframe sees that. The transcript is one signal among several rather than the only one.

The trade is that Wideframe is broader and slightly less specialized. For pure transcript-driven work, Simon Says's polish on the transcript editing surface is hard to match. Wideframe handles transcript queries well but is not specifically optimized for the read-and-select-quotes workflow that drives Simon Says.

Feature Comparison

The comparison divides cleanly along the assembly model. Simon Says leads where transcript polish, multi-language support, and quote-driven assembly are the priority. Wideframe leads where visual search, NLE-native output, and library scale matter.

Real-World Project Fit

Some concrete project types make the choice clearer.

Long-form interview documentary. If the project is built around interview subjects telling their stories and your editorial work is selecting which quotes make the cut, Simon Says is well-suited. The transcript is the editorial surface, and quote selection is the core of the work. Wideframe handles this case but adds the visual search layer for the B-roll and reaction shots that wrap around the interviews. For a project that is mostly interview, Simon Says is more focused. For a project that is interview plus heavy B-roll, Wideframe covers both stages.

Multi-language deposition or oral history. Simon Says's multi-language transcription quality is hard to beat for legal, journalistic, or academic projects that involve subjects speaking in languages other than English. Wideframe's visual search is less relevant for this content -- the value is in the words. Simon Says is the right tool here.

Branded content with on-camera talent and B-roll. Branded videos typically combine on-camera spokesperson footage with B-roll showing the product, environment, or activity. Simon Says handles the spokesperson cuts via transcript. The B-roll has to be selected separately, often manually. Wideframe handles both -- transcript search for the spokesperson, visual search for the B-roll -- in one workflow. For branded content, Wideframe's coverage of both layers is a meaningful advantage.

Narrative film with rough cut assembly. Narrative editing depends on performance, composition, and visual continuity. The transcript captures the dialogue, but the editorial decisions are about which take has the right energy, which cut to which shot, which composition serves the scene. Wideframe's visual semantic search is built for this. Simon Says's transcript-only model leaves most of the relevant signal outside its scope.

Podcast video. If the video is essentially the podcast with two cameras pointed at the hosts, Simon Says is excellent. The story is the words, the visuals are mostly faces talking, and transcript-driven assembly is the most efficient path. Wideframe is overkill for this.

Documentary with deep B-roll library. Documentary work blends interviews with extensive visual storytelling -- archival footage, location shoots, observational moments. The transcript covers a small fraction of the editorial decisions. Wideframe handles the full library. Simon Says covers only the interview component.

Using Them Together

For some teams, the answer is using both tools for the parts each does best. The two are not exclusive.

This pairing is common in documentary teams that need both the transcription depth of Simon Says and the visual search coverage of Wideframe. Each tool stays in its lane, and the combined workflow handles content that neither alone would cover.

Pricing and Pipeline Considerations

Simon Says prices per minute of transcription plus a subscription, which scales with usage. For teams that transcribe heavily, the costs add up. Multi-language transcription costs more than English-only.

Wideframe is a flat $100 per month regardless of footage volume, search volume, or rough cut volume. The pricing reflects that the unit of value is access to the tool, not per-asset processing. For editors with terabyte-scale libraries that they search constantly, the flat pricing is significantly cheaper than per-minute alternatives would be.

The pipeline implications matter too. Simon Says exports interchange formats that work but require import-side cleanup. Wideframe exports native Premiere Pro projects that open directly. For teams that have settled on Premiere as their NLE, the native handoff removes friction. For teams that work across NLEs (some Premiere, some Final Cut, some Resolve), Simon Says's interchange model has slightly broader reach.

Which Fits Your Work?

For most editors whose work involves significant visual storytelling, Wideframe's broader semantic model is the better fit. For editors whose work is genuinely centered on dialogue and quote selection -- especially in multi-language contexts -- Simon Says is purpose-built and excellent.

The honest framing is that these are not direct competitors. They serve overlapping but distinct audiences with different assembly models. Pick based on where your editorial decisions actually live: in the words, or in the picture. For more on the picture-side workflow, see our guide to creating a rough cut in minutes with AI and our comparison of semantic search and transcript search.

Frequently asked questions