Skip to main content
Glama
Hermoso

Hermoso

Official

Edit a video clip

edit_video

Change one named object or element in an existing video clip from a plain-language instruction while motion, timing, framing and cuts stay intact. Works best on 3-15s clips.

Instructions

SWAP AN OBJECT IN A VIDEO ('replace the cup with our can' + the product photo as elements) or EDIT an existing clip from a plain instruction (video-to-video): the motion, timing, framing and cut stay, the named thing changes. 'change only the mug to red', 'make it nighttime', 'restyle it as claymation'. Your words are wrapped so the model keeps everything else identical, changes only what you named and repeats that lock; lighting is kept unless the change needs new light (set lighting to force either); literal:true sends your words as written. One change per call holds best. previewFirstFrame:true edits ONE still first (one image edit; previewAt picks the second) and quotes the clip; nothing else runs until you call again, ideally with previewStill. The reply scores how well the shot held outside the change (free) and flags an edit that touched more than asked. NOT for cuts/trims/end cards (post_edit), a new video (generate_video / render_ad), translating what is SPOKEN (dub_video) or a saved creator's face (recast_motion). Translating or rewriting the on-screen words of a video laid out from a design (a template ad, a custom design) is textChanges with every line at once: one re-make of the design, about 2 credits. Best on 3-15s clips.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "video": "<video>",
      "instruction": "<instruction>"
    },
    "name": "edit_video"
  }
}

Input Schema

NameRequiredDescriptionDefault
videoYesthe source video URL (a render, job result or list_library); a URL, not a local file path; use upload_file for local files
engineNodefault auto: Seedance 2.5 Edit without a real-looking person, else Kling 4.0 Edit
literalNotrue = send the instruction exactly as written, with no preserve/lock wrapper
elementsNoOPTIONAL identity/product grounding (≤4): a creator portrait or the real product photo, so the edit restores the REAL thing. Describe each one in the instruction
get_costNotrue: return the price only; nothing runs or bills
lightingNodefault auto: keep the source light unless the change needs new light (night, a lamp, fire). 'preserve' or 'relight' forces it
faceRouteNo'face_lane' = the user's "I own the rights to this face" for a real person in the source clip (paid plans). Only on the user's say-so, never on your own.
keepAudioNodefault true: keep the source audio; false = silent
previewAtNosecond to preview (default 0); resend with previewStill
referenceNoOPTIONAL image URL that anchors the MATERIAL of what changes (a fabric, a finish, a colour swatch, the real product). Only its surface is used, never its framing or light
instructionYesthe change, in the user’s own words
textChangesNoEVERY change of on-screen words at once ([{from, to}], e.g. each line and its translation). A video laid out from a design is made again with all of them in ONE render, about 2 credits, never a video edit
previewStillNothe still a previewFirstFrame call returned; the clip then matches the changed thing to it
interactionIdNoOPTIONAL: the interactionId an earlier Gemini Omni render or edit returned; the edit continues that clip on the same Omni model. If it cannot run, the video editor edits it and the reply says so.
idempotency_keyNochosen once, resent unchanged on retries: the same key never buys twice
expected_creditsNothe get_cost price; refused unbilled if the live price differs
previewFirstFrameNotrue = edit ONE still of the first frame first and quote the clip; the paid clip does not run

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed3 schema fields changedv0.1.532
    • addedInput schema / properties / expected_credits
      Added value: +{
      +  "description": "the get_cost price; refused unbilled if the live price differs",
      +  "type": "number"
      +}
    • addedInput schema / properties / get_cost
      Added value: +{
      +  "description": "true: return the price only; nothing runs or bills",
      +  "type": "boolean"
      +}
    • addedInput schema / properties / idempotency_key
      Added value: +{
      +  "description": "chosen once, resent unchanged on retries: the same key never buys twice",
      +  "type": "string"
      +}
  2. Changed1 schema field changedv0.1.525
    • changedInput schema / properties / video / description
      Previous value: -"the source video URL (a render, job result or list_library)"New value: +"the source video URL (a render, job result or list_library); a URL, not a local file path; use upload_file for local files"
  3. Changed1 schema field changedv0.1.519
    • addedInput schema / properties / textChanges
      Added value: +{
      +  "description": "EVERY change of on-screen words at once ([{from, to}], e.g. each line and its translation). A video laid out from a design is made again with all of them in ONE render, about 2 credits, never a video edit",
      +  "items": {
      +    "properties": {
      +      "from": {
      +        "type": "string"
      +      },
      +      "to": {
      +        "type": "string"
      +      }
      +    },
      +    "required": [
      +      "from",
      +      "to"
      +    ],
      +    "type": "object"
      +  },
      +  "type": "array"
      +}
  4. Changed1 schema field changedv0.1.467
    • changedInput schema / properties / engine / description
      Previous value: -"default auto: Seedance 2.5 Edit without a real-looking person, else Kling O3 Edit"New value: +"default auto: Seedance 2.5 Edit without a real-looking person, else Kling 4.0 Edit"
  5. Changed2 schema fields changedv0.1.374
    • addedInput schema / properties / engine
      Added value: +{
      +  "description": "default auto: Seedance 2.5 Edit without a real-looking person, else Kling O3 Edit",
      +  "enum": [
      +    "auto",
      +    "seedance",
      +    "kling"
      +  ],
      +  "type": "string"
      +}
    • addedInput schema / properties / faceRoute
      Added value: +{
      +  "description": "'face_lane' = the user's \"I own the rights to this face\" for a real person in the source clip (paid plans). Only on the user's say-so, never on your own.",
      +  "enum": [
      +    "face_lane"
      +  ],
      +  "type": "string"
      +}
  6. Changed11 schema fields changedv0.1.320
    • changedInput schema / properties / elements / description
      Previous value: -"OPTIONAL identity/product grounding (≤4): a creator portrait or the real product photo, so the edit restores the REAL thing instead of re-inventing it. Describe each one in the instruction. Leave out for a plain restyle"New value: +"OPTIONAL identity/product grounding (≤4): a creator portrait or the real product photo, so the edit restores the REAL thing. Describe each one in the instruction"
    • changedInput schema / properties / instruction / description
      Previous value: -"the exact transformation to apply, in the user’s own words"New value: +"the change, in the user’s own words"
    • changedInput schema / properties / interactionId / description
      Previous value: -"OPTIONAL: the interactionId an earlier Gemini Omni render or edit returned. The edit then continues that clip on the SAME Omni model from its own stored context (identity-true, no re-upload, usually cheaper). If that edit cannot run, the clip is edited by the video editor instead and the reply says so."New value: +"OPTIONAL: the interactionId an earlier Gemini Omni render or edit returned; the edit continues that clip on the same Omni model. If it cannot run, the video editor edits it and the reply says so."
    • changedInput schema / properties / keepAudio / description
      Previous value: -"default true — keep the source clip’s audio track. Set false to return the edit silent"New value: +"default true: keep the source audio; false = silent"
    • addedInput schema / properties / lighting
      Added value: +{
      +  "description": "default auto: keep the source light unless the change needs new light (night, a lamp, fire). 'preserve' or 'relight' forces it",
      +  "enum": [
      +    "auto",
      +    "preserve",
      +    "relight"
      +  ],
      +  "type": "string"
      +}
    • addedInput schema / properties / literal
      Added value: +{
      +  "description": "true = send the instruction exactly as written, with no preserve/lock wrapper",
      +  "type": "boolean"
      +}
    • addedInput schema / properties / previewAt
      Added value: +{
      +  "description": "second to preview (default 0); resend with previewStill",
      +  "type": "number"
      +}
    • addedInput schema / properties / previewFirstFrame
      Added value: +{
      +  "description": "true = edit ONE still of the first frame first and quote the clip; the paid clip does not run",
      +  "type": "boolean"
      +}
    • addedInput schema / properties / previewStill
      Added value: +{
      +  "description": "the still a previewFirstFrame call returned; the clip then matches the changed thing to it",
      +  "type": "string"
      +}
    • addedInput schema / properties / reference
      Added value: +{
      +  "description": "OPTIONAL image URL that anchors the MATERIAL of what changes (a fabric, a finish, a colour swatch, the real product). Only its surface is used, never its framing or light",
      +  "type": "string"
      +}
    • changedInput schema / properties / video / description
      Previous value: -"the source video URL (from a previous render, a job result, or list_library)"New value: +"the source video URL (a render, job result or list_library)"
  7. Changed1 schema field changedv0.1.252
    • addedInput schema / properties / interactionId
      Added value: +{
      +  "description": "OPTIONAL: the interactionId an earlier Gemini Omni render or edit returned. The edit then continues that clip on the SAME Omni model from its own stored context (identity-true, no re-upload, usually cheaper). If that edit cannot run, the clip is edited by the video editor instead and the reply says so.",
      +  "type": "string"
      +}
  8. Addedv0.1.161

TDQS

A4.6/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only declare read/write and openness; the description adds substantial invisible behavior: the preserve/lock wrapper, lighting defaults and how to force them, that the reply includes a free hold-quality score and an over-edit flag, that previewFirstFrame only quotes and bills nothing until re-called, and roughly 2 credits for design textChanges. This is well beyond what the structured fields convey.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Purpose and exclusions are front-loaded, but the middle is a run-on paragraph mixing wrapper mechanics, preview flow, cost, and reply semantics with idiosyncratic shorthand ('repeats that lock', 'quotes the clip') that an agent must decode. Several clauses duplicate schema descriptions, and the density makes it easy to miss the preview workflow.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With 17 parameters and no output schema, the description does the necessary work: it explains the preview/paid two-step, idempotency intent, cost checking, and what the response contains. Gaps are minor — engine selection trade-offs and elements-count behavior are left mostly to schema text — but nothing critical to a correct call is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the schema already explains each field; the description nonetheless adds cross-parameter meaning the schema lacks — previewStill pairing with a prior previewFirstFrame call and previewAt, expected_credits being refused against the get_cost price, and textChanges being one bundled re-make rather than a video edit. That relational guidance is genuinely additive, though some phrasing simply restates schema text (literal, lighting).

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Opens with a specific verb+resource split (object swap vs. video-to-video instruction edit) and gives concrete example instructions. It explicitly names what it is NOT for and routes to post_edit, generate_video/render_ad, dub_video and recast_motion, so an agent can separate it from every sibling that matters.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

States when to use (swap/edit from a plain instruction), when not to (cuts, trims, end cards, new video, spoken translation, recast), and the operant limits: one change per call, best on 3-15s clips, preview-first recommended. Alternatives are named with the condition that selects each.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server

Other Tools