MCP App Store
by KeenEthics
NextClip AI Video Editor icon

NextClip AI Video Editor MCP Server for ChatGPT

By NextClip

Overview

NextClip edits talking-head videos by working directly from the transcript. It can find and remove filler words, false starts, and repeated takes, correct misheard words, add styled captions, and suggest moments where extra footage (B-roll) would strengthen the video, then add it for you. Start by attaching a video or linking to a public MP4 or MOV file.

How to connect

  1. ChatGPT: Open the NextClip AI Video Editor listing in the ChatGPT app directory and choose Connect. ChatGPT listing NextClip AI Video Editor exposes 16 tools here.

Tools

add_brolls

ChatGPT
Adds b-roll clips chosen from search_broll_footage. For each one give the moment (sentence_id from get_transcript, or from_word_id and to_word_id) and the Pexels media_id and media_type. Never pass URLs. B-roll covers the spoken words of that moment, up to max_seconds (default 5), and is shortened if the clip is shorter. Overlapping b-roll and parts of the video that were cut are skipped, and the reasons are returned. Add at most 8 per call. Do your cuts first, because b-roll is timed to the edited video. Avoid covering the first 2 seconds. theme: Simple-broll is full screen (default). The other themes also change the layout of the main video (half-broll, broll-in-center, video-on-top). Tell the user what was added and that they can review it in the editor.

analyze_cuts

ChatGPT
Read only quick scan that counts filler words (um, uh), stutters and long pauses and estimates the time saved. It does not return the transcript. Use get_transcript to read the speech and find retakes, false starts and irrelevant sentences. Show the user the numbers and ask before calling apply_cuts with the same options.

apply_captions

ChatGPT
Adds animated captions to the project using a caption style, and saves them to the draft. Call list_caption_styles first if the user wants to pick a style, and pass the exact id. Omit style_id to keep the current style (or the default for a new project). Run this after cutting, so captions match the edited timeline. Calling it again with a different style replaces the captions.

apply_cuts

ChatGPT
Removes or restores parts of the video and saves the draft. Read the transcript with get_transcript first, decide what to cut, and (unless the user already told you exactly what to remove) confirm with the user. action "remove": you can combine any of: sentence_ids (from get_transcript, the easiest way to drop whole sentences), word_ranges (word ids, for single words or part of a sentence), time_ranges (milliseconds on the ORIGINAL video, for pauses or non-speech), and the automatic helpers remove_fillers / remove_repeats / silence_threshold_ms. action "restore": pass sentence_ids or word_ranges to bring parts back, or restore_all=true to undo every cut. Cuts are added to existing ones. Use ids from the latest get_transcript call only, and call it again after correct_transcript. Do cutting before adding b-roll or text, because visuals are timed to the edited video. Share the new length and the editor link afterwards.

correct_transcript

ChatGPT
Fixes misspelled or misheard words in the transcript. edits: change specific words by id (word ids come from get_transcript). new_text may be more than one word. replacements: change every occurrence of one single word, such as a brand or name (case insensitive unless match_case is true); punctuation around the word is kept. This only changes the text, not the audio or timing, so the spoken audio is unchanged. Existing captions are regenerated. Tell the user exactly what changed.

create_project

ChatGPT
Creates a NextClip project from a video and starts transcribing it. If the user attached or uploaded a video in the chat, pass it as video_file. Otherwise pass video_url, a public https link directly to an .mp4 or .mov file (YouTube links are not supported). Provide exactly one of the two. Call get_user_plan first to check credits and limits. This spends 10 credit per project, so mention the cost if it is not obvious. After this returns, call get_project to wait for the transcript.

export_project

ChatGPT
Renders the project (cuts, captions and any visuals) into a final MP4. Exporting is only available on paid plans. If the project has no captions yet, default captions are added automatically. resolution is optional (for example 1080p); the options depend on the aspect ratio and the plan, and an invalid value returns the list. aspect_ratio is optional: omit it to match the source video, or pass one of the supported ratios (for example 9:16 for a vertical short). Rendering is asynchronous and takes a few minutes. Afterwards call get_export_status when the user asks whether it is ready. Starting an export replaces any earlier exported file for this project.

get_export_status

ChatGPT
Checks whether the exported MP4 for a project is ready and returns its URL. Only call this after export_project.

get_project

ChatGPT
Returns a project's status (processing, ready, failed), length, how much has been cut, and the link to open it in the NextClip editor. If the project is still processing this waits up to wait_seconds, so call it again if it still says processing. Omit project_id to list the user's 5 most recent projects. Never guess a project_id.

get_transcript

ChatGPT
Returns the project's transcript so you can read, quote, summarize or edit it. view "sentences" (default): each sentence has id, startMs, endMs, firstWordId, lastWordId, text, pauseAfterMs and removed (none, partial or all, meaning already cut). view "words": word level detail for a word id range, as arrays [wordId, text, startMs, endMs, removed(0 or 1)]. Use it to find the exact word to fix or cut. Times are milliseconds on the ORIGINAL video. Page with from_sentence if nextFromSentence is not null. Use this to find retakes (the same thing said twice, keep the best take), false starts, tangents and filler. Never invent text that is not returned here.

get_user_plan

ChatGPT
Returns the user's current subscription plan, available credits, and which features are available to them. Call this at the start of each session and before generate_clips, export_clip, or schedule_clip. 1 credit = 1 minute of video processed.

get_visuals

ChatGPT
Lists everything visual in the draft with ids, times on the edited video (seconds) and the words spoken under each one. Use it before replacing or removing visuals, and to see what is already there before adding more.

list_caption_styles

ChatGPT
Returns the available caption styles with their ids. Call this before apply_captions when the user wants to choose or change a style. Never guess a style id.

remove_visuals

ChatGPT
Removes b-roll, text or overlay items by id (ids from get_visuals or add_brolls). Up to 50 at once. This only removes them from the draft.

replace_broll

ChatGPT
Replaces the footage of an existing b-roll (id from get_visuals or add_brolls) with another Pexels clip, keeping its timing. Use the candidates from an earlier search_broll_footage call, or search again. Optionally change the theme.

search_broll_footage

ChatGPT
Searches Pexels stock footage and returns candidates for each search: id, description, durationSec and a pageUrl the user can open. Run one search per b-roll moment (up to 8 at once) with 3 to 5 concrete, visual words, not abstract ideas (good: 'office team celebrating', bad: 'success'). Set min_seconds to the length of the moment so short clips are filtered out. You cannot see the footage, only the descriptions, so choose the best description and offer the user the alternatives. Nothing is added until you call add_brolls with the chosen ids. Use video unless the user asks for photos.

Capabilities

Writes

Example Prompts

Click any prompt to copy it.

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/nextclip-ai-video-editor"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/nextclip-ai-video-editor)