Tuney
Server Details
Royalty-free music fitted to your length or video: find, generate and adjust tracks, score videos.
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- OAuth
- Works in Glama
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- Repository
- library-x/tuney_mcp
- GitHub Stars
- 0
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 14 tools
Most tools target distinct resources/actions (find vs generate vs adjust vs get vs download), and the video tools are split sensibly by purpose. The one fuzzy boundary is score_video with render=true versus render_video (which also places a track under a video), plus the intentional get_task/wait_for_task pair; descriptions do resolve both, so misselection risk is low.
Every tool is snake_case verb_noun (adjust_track, find_tracks, generate_track, get_track, score_video, render_video, wait_for_task). No mixed conventions or vague single-word tool names.
14 tools is squarely in the well-scoped range for a music/video generation service with background jobs. Each tool maps to a real workflow step (discover, generate, edit, score, download, task polling) with no filler.
The surface covers the full lifecycle: style discovery, search, generation, re-editing, retrieval, credit check, download, video import/analysis/scoring/rendering, and async task handling. The main gap is no way to list the user's own tracks or history — a cue_id must be remembered from a prior result — which is workable but slightly limiting.
Available Tools
14 toolsadjust_trackAdjust a trackAInspect
Re-render a track with changes, keeping it musically coherent.
- length_seconds: fit the track to an exact duration (the engine recomposes the arrangement).
- turning_point_seconds: add a turning point at that second, where the music changes character
(new arrangement after the point, an audible transition). Combine with length_seconds to
lengthen a track and place the turning point in one go. If the track already has a turning
point, this moves it.
- ending: how the track ends, "fade_out" or "short" (a stab on the last beat). The track keeps
its length (or the length_seconds you give) within a quarter of a second either way.
- time_stretch: tempo factor, for example 0.9 slower or 1.1 faster.
- pitch_shift: semitones, for example -2 or 3.
- mute / unmute: stems by name: melody, chords, bass, drums.
- regenerate: stems to re-pick musical material for (melody, chords, bass, drums); all four
give a fresh take on the same brief.
Runs in the background: returns a task_id for wait_for_task.
Inputs: cue_id (required), mute, ending, unmute, +5 more. Returns: object
analyze_videoAnalyze a videoARead-onlyIdempotentInspect
Show what Tuney sees in a video: picture cuts with their strength, where people speak, and a sonic-logo guess.
Optional: score_video analyses the video by itself. Use this to discuss the edit with the user
first, for example to pick the turning point or the moment the music should end.
Inputs: video_id (required). Returns: object
download_trackDownload a trackAInspect
Get a download link for the final file. This spends Tuney credits (mp3, wav, or a zip of stems).
Returns download_url right away when the file is cached, otherwise a task_id to pass to
wait_for_task with kind="download". Links expire, so share them promptly. download_url is on
the Tuney API host; download_direct_url is the same file in Cloud Storage.
For format="stems" the link is a zip with one mp3 per part; the result's `contents` map
gives the readable name of each file (melody, chords, bass, drums, swoosh fx, impact fx).
To show stems to the user: fetch the zip, unzip it, rename the files with that map and
attach them. Never download the same track twice.
Inputs: cue_id (required), format. Returns: object
find_tracksFind ready-made tracksARead-onlyIdempotentInspect
Find ready-made Tuney tracks that match a brief. Results are instant and previews are free.
Give either a free-text prompt (for example "warm acoustic guitar for a travel vlog intro") or
exact genre, mood and tempo values from list_styles. Each result has a cue_id and a preview
URL. Pass the returned search_id to get more results for the same brief. Use adjust_track
to change length or key, and download_track for the final file (that costs credits).
Inputs: mood, genre, limit, tempo, +2 more. Returns: object
generate_trackGenerate a new trackAInspect
Compose a brand-new track from Tuney's hand-made musical elements, fitted to an exact length.
Use exact genre, mood and tempo values from list_styles, or leave them empty for any style.
Generation runs in the background: this returns a task_id, so call wait_for_task next.
Previews are free; downloading the final file costs credits.
Inputs: mood, build, genre, tempo, +1 more. Returns: object
get_accountShow the signed-in accountARead-onlyIdempotentInspect
Show which Tuney account is connected and its credit balance, with a link to buy more.
Inputs: no named parameters. Returns: object
get_taskCheck a background taskARead-onlyIdempotentInspect
Check a background task once: a generation, a download, or a video import, analysis, scoring or render.
Prefer wait_for_task, which polls for you.
Inputs: task_id (required), kind. Returns: object
get_trackGet a trackARead-onlyIdempotentInspect
Get a track by cue_id: name, preview URL, length, bpm, key, style and sections.
Inputs: cue_id (required). Returns: object
get_video_upload_linkGet a link to upload a videoAInspect
Create a one-time link for sending a video file (MP4, MOV or M4V) to Tuney for scoring.
Use this when the video is a local file, for example one the user attached or one in the
working directory. Send the file with the curl command in `example` (HTTP PUT of the raw
file); the response contains the video_id to pass to score_video. If the video is already
online, use import_video with its link instead.
Inputs: filename. Returns: object
import_videoImport a video from a linkAInspect
Bring a video into Tuney from a public https link that returns the file itself (not a web page).
Returns the video_id to pass to score_video, analyze_video or render_video. Videos up to
200 MB and 10 minutes. For a local file, use get_video_upload_link instead.
Inputs: url (required). Returns: object
list_stylesList music stylesARead-onlyIdempotentInspect
List the genres, moods and tempos Tuney can find or generate music for.
Call this before find_tracks or generate_track when the user names a style, and pass the
exact values returned here.
Inputs: no named parameters. Returns: object
render_videoRender a video with a trackAInspect
Put one of the user's tracks under a video, for example after adjust_track changed the scored track.
duck lowers the music while people speak. music_volume and video_volume are gains from 0 to 2
(1 leaves the level unchanged). Returns a link to the rendered MP4.
Inputs: cue_id (required), video_id (required), duck, music_volume, +1 more. Returns: object
score_videoScore a videoAInspect
Compose music to a video and, by default, return the video with the music under it.
Tuney finds the picture cuts, speech and sonic logo, then arranges a track to them: the music
ends where the logo starts (or at the end), the turning point lands on the strongest cut in
the middle, hits land on other strong cuts outside speech, and the music is lowered while
people speak in the rendered video.
- prompt: describe the music ("upbeat pop for a sneaker ad"); or cue_id: one of the user's tracks.
- ending: "fade" lands exactly on the length; "short" ends on a stab.
- turning_point_seconds / no_turning_point: override the automatic turning point.
- music_end_seconds / logo_start_seconds: where the music must stop (a sonic logo starts).
- hits: sound effects on strong cuts; render: also return the scored video.
Scoring takes about a minute; if the result says timed_out, call wait_for_task with the
task_id and kind="score_video". The result has the track (preview_url), the plan, and the
video link when rendered. Previews and renders are free; downloading the track costs credits.
Inputs: video_id (required), hits, cue_id, ending, +6 more. Returns: object
wait_for_taskWait for a background taskARead-onlyIdempotentInspect
Wait for a background task to finish, reporting progress along the way.
kind is the one the starting tool returned: generate, download, import_video, analyze_video,
score_video or render_video. Returns the finished track (with preview_url), the download_url,
or the video result. If it is still running when the timeout passes, the result says so:
call wait_for_task again with the same task_id and kind.
Inputs: task_id (required), kind, timeout_seconds. Returns: object
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Licensed soundtracks and sound effects matched to your video. Commercial use OK.
Human-made production music for sync. Search by brief or reference, preview, and score to picture.
Search 2M+ licensed production-music tracks, find similar ones by track or audio, share pick lists.
Generate, edit and stream royalty-free music, or search a licensed catalogue.
Related MCP Servers
- FlicenseNot gradedqualityAmaintenanceEnables agents to analyze video footage for mood, pace, speech, and scene cuts, then return a ranked shortlist of license-safe royalty-free or Creative Commons tracks with the license printed on every row. Also supplies a 12–20s hook window, beat-aligned cut hints, and an ffmpeg ducking spec for mixing.-
- AlicenseAqualityBmaintenanceEnables AI clients to compose original game music across 19 styles plus 32 sound effects and 11 looping ambiences, rendered deterministically to WAV (and optionally MIDI) from a seed and musical parameters. Everything is synthesized in pure Node.js with no dependencies, producing seamless loops, mono or multi-rate output, and reproducible renders for game soundtracks, menus, and environmental beds.4509 npmMIT

AINSOF MCP Serverofficial
AlicenseNot gradedqualityBmaintenanceEnables AI assistants to search a curated production-music catalogue by brief or reference link, listen to full previews, and score videos with synced music, plus retrieve stems, versions, and cue sheets for licensing.MIT- FlicenseNot gradedqualityAmaintenanceEnables agents to ingest video, analyze mood, pace, speech, and scenes, and return ranked license-safe BGM matches with license and attribution details. It also provides 12–20s hook windows, beat-synced cut suggestions, and ffmpeg ducking or mix specs for editing workflows.-
Glama MCP Gateway
Add one secure layer between your agents and this server.