MiniMax MCP Server by useapi.net (third-party)
Server Details
MiniMax (Hailuo) video, images, music, speech and voice cloning, templates and the AI agent
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 38 tools
Each tool targets a distinct resource+action (create/get/list/delete per modality, plus upload, agent, account tools), and the descriptions explicitly cross-reference near-neighbours — e.g. minimax_cancel_video vs minimax_stop_tracking vs minimax_delete_video, and minimax_get_account vs minimax_get_features vs minimax_list_accounts are spelled out as different operations. The trickiest pairs (minimax_list_running_jobs vs minimax_list_videos/minimax_list_images, minimax_list_agent_jobs vs minimax_get_agent_job) are clearly delimited in their text.
Nearly everything follows a strict minimax_verb_noun snake_case pattern (minimax_create_video, minimax_get_music, minimax_delete_file, minimax_upload_voice_sample). Minor deviations: minimax_stop_tracking has no clear noun object, and job naming flips between singular (minimax_get_agent_job) and plural (minimax_list_agent_jobs), but the scheme remains predictable.
38 tools is heavy for any single server; while the domain (video, image, music, speech, agent, files, accounts) is genuinely broad, four parallel CRUD families plus two upload families and separate voice/config/quota tools create a large surface. Some consolidation (e.g. unified delete/upload, or per-modality grouping) would reduce cognitive load.
Coverage is strong: create/get/list/delete for video, image, music and speech, plus uploads, voice cloning/quota/config, agent jobs and account/feature inspection. Gaps are minor — no cancel for image, music or speech jobs (only video has minimax_cancel_video), no update/rename operations, and account creation is out of band on a setup page.
Publisher details
- Operator
- useapi.net
- Operator website
- https://useapi.net
- Vendor relationship
- Independent
- Documentation
- https://useapi.net/docs/api-minimax-v1/mcp
- Privacy policy
- Not provided
- Trust center
- https://useapi.net/docs/legal
- Restrictions
- Paid useapi.net subscription ($15/month flat). Requires your own MiniMax (Hailuo AI or MiniMax Audio) account. · Publisher source
Available Tools
38 toolsminimax_cancel_videoCancel a running videoADestructiveIdempotentInspect
Ask MiniMax to cancel a video that is still generating, by videoId (from minimax_create_video or minimax_list_running_jobs). An empty success means MiniMax accepted the request; read minimax_get_video for the video's final status. To only stop this API tracking a job use minimax_stop_tracking; to remove a finished video use minimax_delete_video. Cannot be undone. POST /videos/cancel.
Inputs: videoId (required).
minimax_clone_voiceClone a voiceAInspect
Build a new voice from 1-10 samples uploaded with minimax_upload_voice_sample (same account). Takes a cloned voice slot (minimax_get_voice_quota shows them; out of slots is an error). Returns the voice_id to pass to minimax_create_speech, voice_name, tag_list and a sample_audio URL. Cloned voices are listed by minimax_list_voices with is_system false. POST /speech/clone-voice.
Inputs: files (required), voice_name (required), language_tag (required), need_noise_reduction.
minimax_create_imageGenerate imagesAInspect
Generate 1-4 images on hailuoai.video with image-01, MiniMax H3 Image (h3-image, -highspeed), Nano Banana 2 / Pro, GPT Image 1.5 / 2, Seedream 4.5 / 5.0 or Midjourney V7 / Niji 7 (always 4), from a prompt and up to 16 reference images (referenceFileIdN, cited as @ImageN on models that support it) or a character image (characterFileId, image-01). Uses MiniMax credits per image. Waits up to ~50 s and returns each image's URL and fileID, the first as an image; else the imageIds for minimax_get_image. POST /images/create.
Inputs: prompt (required), wait, model, account, +24 more.
minimax_create_musicCreate a songAInspect
Generate 1-3 songs with MiniMax Music (music-3.0, music-2.6) from a style prompt, your own lyrics, or a reference track (a cover); for spoken text use minimax_create_speech. Supply prompt, lyrics, referenceMusicId or a combination. One song generation per MiniMax account at a time (429 while one runs, including songs started on minimax.io); uses the account's MiniMax Audio credits, not its Hailuo video credits (412 without them). Waits up to ~50 s and returns each track's musicId, status and, once completed, audio_url; else call minimax_get_music. POST /music/create.
Inputs: wait, model, title, lyrics, +7 more.
minimax_create_speechCreate speechAInspect
Convert text to speech with a MiniMax voice (speech-2.8-hd, speech-2.8-turbo, speech-2.6-hd, speech-2.6-turbo, speech-2.5-hd-preview, speech-2.5-turbo-preview, speech-02-hd, speech-02-turbo, speech-01-hd, speech-01-turbo; default speech-2.8-hd), up to 5000 characters on -hd models and 10000 on -turbo. voice_id comes from minimax_list_voices (built-in or cloned with minimax_clone_voice); emotion and language_boost values from minimax_get_speech_config. Billed per character in the account's MiniMax Audio credits (the subscription music uses, not Hailuo video credits; 412 without them); too many at once on an account answers 429. Waits up to ~50 s and returns the audioId and, once done, audio_url; else call minimax_get_speech. POST /speech/create.
Inputs: text (required), voice_id (required), vol, wait, +15 more.
minimax_create_template_videoGenerate a template videoAInspect
Generate a video from a hailuoai.video agent template (templateId from minimax_list_agent_templates), with the prompt and / or image the template requires (check it with minimax_list_agent_templates and templateId). Uses the account's MiniMax credits (the template's credits). Waits up to ~50 s, then returns the videoId for minimax_get_video; free-form generation is minimax_create_video. POST /videos/agent-create.
Inputs: templateId (required), wait, fileID, prompt, +5 more.
minimax_create_videoGenerate a videoAInspect
Generate a video on hailuoai.video with Hailuo 3.0, 02 / 2.3, Hailuo 01, Veo 3.1 (incl. subject reference S2V) or Seedance 2.0 / Fast / Mini, from text, a first frame (fileID), an end frame, subject images (fileID2-14 on S2V) or omni references cited in the prompt as @Image1-9, @Video1-3, @Audio1-3 (Seedance, Hailuo-3.0). Hailuo 02 / 2.3 and Veo set size and length with options; Seedance and Hailuo-3.0 with resolution and duration. Uses the account's MiniMax credits (minimax_get_features). Waits up to ~50 s, then returns the videoId for minimax_get_video. POST /videos/create.
Inputs: wait, model, fileID, prompt, +29 more.
minimax_delete_fileDelete an uploaded fileADestructiveIdempotentInspect
Delete one uploaded reference file from its account by fileID (from minimax_upload_file or minimax_list_files); generations that already used it are not affected. Answers success also for a file that does not exist or was already deleted. Cannot be undone. DELETE /files/{fileID}.
Inputs: fileID (required).
minimax_delete_imageDelete an imageADestructiveIdempotentInspect
Delete one image from its account's hailuoai.video history by imageId: it leaves minimax_list_images but minimax_get_image can still read it (MiniMax soft-deletes images). Answers success also for an image that does not exist or was already deleted. Cannot be undone. DELETE /images/{imageId}.
Inputs: imageId (required).
minimax_delete_musicDelete a songADestructiveIdempotentInspect
Delete one song or reference track from its MiniMax account by musicId; its audio_url stops working, so download it first. A song still rendering cannot be deleted (409): wait until minimax_get_music shows statusFinal true. The account comes from the musicId. Cannot be undone. DELETE /music/{musicId}.
Inputs: musicId (required).
minimax_delete_speechDelete speechADestructiveIdempotentInspect
Delete one speech recording from its MiniMax account by audioId or audio_id; it leaves the account and minimax_list_speech. Answers success for a recording that does not exist or is already deleted. A generation still running cannot be deleted (400): wait until minimax_get_speech shows it final, then delete it. minimax_stop_tracking does not stop it and leaves this audioId undeletable; then delete the finished recording by its audio_id from minimax_list_speech. Cannot be undone. DELETE /speech/{audio_id}.
Inputs: audio_id (required).
minimax_delete_videoDelete a videoADestructiveIdempotentInspect
Delete one video from its account's hailuoai.video history by videoId. Answers success also for a video that does not exist or was already deleted. To stop a video that is still generating use minimax_cancel_video. Cannot be undone. DELETE /videos/{videoId}.
Inputs: videoId (required).
minimax_delete_voiceDelete a cloned voiceADestructiveIdempotentInspect
Delete one cloned voice by its voice_id (from minimax_clone_voice or minimax_list_voices with is_system false), freeing its slot. The account comes from the voice_id. Answers success for a voice that does not exist or is already deleted. Cannot be undone. POST /speech/delete-voice.
Inputs: voice_id (required).
minimax_get_accountGet one accountARead-onlyIdempotentInspect
One stored MiniMax account by id: host, maxJobs, supported features (supportVideo, supportAudio, supportMusic, supportAgent, supportChat), session issue and expiry times, and the stored account error if MiniMax put it on hold. No call to MiniMax; use minimax_get_features for live credits, minimax_list_accounts for all accounts. GET /accounts/{account}.
Inputs: account (required).
minimax_get_agent_jobGet an agent jobARead-onlyIdempotentInspect
One agent job by jobId (from minimax_run_agent or minimax_list_agent_jobs): status (started, completed, failed), the agent's message and, once completed, each file it made with fileName, type, fileUrl and no_watermark_url (videos); a failure's reason. wait: true waits up to ~50 s. Jobs are kept 31 days. GET /agent/{jobId}.
Inputs: jobId (required), wait.
minimax_get_featuresGet credits and planARead-onlyIdempotentInspect
Ask MiniMax live for one account's video/image credits (totalCredits), plan (memberName, isPaidSubscription) and per-model settings (modelControl, as MiniMax reports them), plus the minimax.io audio credits for music and speech (audio.total_credit). Use before a generation to check credits, or after a 412; minimax_get_account shows stored settings without calling MiniMax. GET /features.
Inputs: account.
minimax_get_imageGet an imageARead-onlyIdempotentInspect
One image by imageId (from minimax_create_image or minimax_list_images): status label (pending, queued, processing, completed, failed, moderated) and, once completed, its url (videoURL), downloadURL (no watermark, paid accounts), size and fileID (reusable as a reference), with the image itself as an image block. wait: true waits up to ~50 s. Deleted images can still be read here. GET /images/{imageId}.
Inputs: imageId (required), wait.
minimax_get_musicGet a songARead-onlyIdempotentInspect
One song from minimax_create_music, or a reference from minimax_upload_music_reference, by musicId: statusLabel (pending, processing, completed, failed, moderated), statusFinal and, once completed, audio_url (MP3), duration, title, lyrics and model; typeLabel tells a song from a reference, referenceMusicId marks a cover. wait: true waits up to ~50 s for an unfinished song. minimax_list_music lists many. GET /music/{musicId}.
Inputs: musicId (required), wait.
minimax_get_speechGet speechARead-onlyIdempotentInspect
One speech recording by audioId (from minimax_create_speech) or audio_id (from minimax_list_speech): statusLabel (pending, processing, completed, failed), statusFinal and, once completed, audio_url (MP3), text, model, voice_info, settings and cost_credit (the credits MiniMax charged); a failure's code and error. wait: true waits up to ~50 s for an unfinished one. An audioId from minimax_create_speech resolves for 7 days; after that find the recording with minimax_list_speech. GET /speech/{audio_id}.
Inputs: audio_id (required), wait.
minimax_get_speech_configGet speech settingsARead-onlyIdempotentInspect
The vocabularies speech accepts on an account: emotion values (t2a_emotion, for emotion in minimax_create_speech), and the voice tags by language, gender, age and accent (tag_name values for tag_list in minimax_list_voices; the language names also serve as language_boost and as language_tag in minimax_clone_voice). Read it before choosing those inputs. GET /speech/config.
Inputs: account.
minimax_get_videoGet a videoARead-onlyIdempotentInspect
One video by videoId (from minimax_create_video, minimax_create_template_video or minimax_list_videos): status label (pending, queued, processing, completed, failed, moderated), percent and message while it runs, and once completed width, height, videoURL, downloadURL (no watermark, paid accounts) and coverURL. wait: true waits up to ~50 s. A 404 usually means the video was moderated and removed. GET /videos/{videoId}.
Inputs: videoId (required), wait.
minimax_get_voice_quotaGet cloned voice slotsARead-onlyIdempotentInspect
How many cloned voice slots an account holds (total) and how many are in use (used), plus free clone previews left ("inf" when unlimited). Check it before minimax_clone_voice; minimax_delete_voice frees a slot. The count is advisory: MiniMax can let used exceed total, and the clone itself gives the final refusal. GET /speech/equity.
Inputs: account.
minimax_list_accountsList MiniMax accountsARead-onlyIdempotentInspect
List the MiniMax (hailuoai.video / minimax.io) accounts added to this useapi.net API token, as stored: account id, host, maxJobs, the features it supports (video, audio, music, agent), session expiry and any stored account error. No call to MiniMax; use minimax_get_features for live credits and plan. Accounts are added on the setup page, not here. GET /accounts.
Inputs: no named parameters.
minimax_list_agent_jobsList running agent jobsARead-onlyIdempotentInspect
The agent jobs running now for this API token, grouped by account, with each jobId and its elapsed time (mm:ss), plus the accounts that can run the agent. The first check when minimax_run_agent answers 429; finished jobs are read with minimax_get_agent_job. Video and image jobs are listed by minimax_list_running_jobs. GET /agent/jobs.
Inputs: no named parameters.
minimax_list_agent_templatesList video templatesARead-onlyIdempotentInspect
Browse hailuoai.video agent video templates for minimax_create_template_video: each template's id (the templateId), title and preview video. With templateId, one template's details instead: credits (the price its page shows), requirePrompt, requireImage, the prompt placeholder and inputs. Pages of limit items (default 30); pass the last id as lastId for the next page while hasNext is true. GET /videos/agent-templates.
Inputs: limit, lastId, account, templateId.
minimax_list_charactersList charactersARead-onlyIdempotentInspect
List the characters saved on one hailuoai.video account: characterID, characterType and the character image (file.fileID, cdnUrl). Use a file.fileID as fileID in minimax_create_video (e.g. S2V-01 subject reference) or minimax_create_template_video. Pages: page (default 1) and pageSize (default 50). Uploaded files are listed by minimax_list_files. GET /videos/characters.
Inputs: page, account, pageSize.
minimax_list_filesList uploaded filesARead-onlyIdempotentInspect
List the reference files uploaded to one account (with minimax_upload_file, POST /files or the MiniMax web app): file_id (the fileID for the create tools), file_name, file_type, cdn_url and oss_path (the ossPath minimax_create_template_video needs). limit sets how many (default 50). Characters made in the web app are listed by minimax_list_characters. GET /files.
Inputs: limit, account.
minimax_list_imagesList imagesARead-onlyIdempotentInspect
List one account's images on hailuoai.video, including running ones and those made in the web app: imageId, statusLabel, URLs, fileID and prompt (desc). Pages of limit items (default 30); pass the last imageId as lastImageId for the next page while hasNext is true. Deleted images are not listed. GET /images.
Inputs: limit, account, lastImageId.
minimax_list_musicList songsARead-onlyIdempotentInspect
List one MiniMax account's music, newest first, including songs made on minimax.io and tracks still rendering: musicId, title, statusLabel, duration, audio_url, typeLabel (song or reference) and referenceMusicId for covers. type picks songs (default), references or all. Page with lastMusicId (the last musicId of the previous page). minimax_get_music reads one song. GET /music.
Inputs: type, limit, account, favorite, +1 more.
minimax_list_running_jobsList running jobsARead-onlyIdempotentInspect
The videos, images, music and speech this API is tracking now (videoId / imageId / musicId / audioId, started, elapsed mm:ss) and the accounts with a free video slot (account, maxJobs, executing, available). The first check when a create tool answers 429 (accounts at maxJobs); everything on an account, web app included, is in minimax_list_videos / minimax_list_images. GET /scheduler and GET /scheduler/available.
Inputs: no named parameters.
minimax_list_speechList speechARead-onlyIdempotentInspect
List one MiniMax account's speech recordings, newest first (made with the API or on minimax.io): audio_id, audio_title, text, voice_name, audio_url and whether SRT or WAV renditions exist, a page at a time; when has_more is true pass the next page number. cost_credit, model and voice settings are only on minimax_get_speech. GET /speech.
Inputs: page, account, page_size.
minimax_list_videosList videosBRead-onlyIdempotentInspect
List one account's videos on hailuoai.video, including running ones and those made in the web app: videoId, statusLabel, statusFinal, URLs and prompt (desc). Pages of limit items (default 30); pass the last videoId as lastVideoId for the next page while hasNext is true. Only what this API is tracking now: minimax_list_running_jobs. GET /videos.
Inputs: limit, account, lastVideoId.
minimax_list_voicesList voicesARead-onlyIdempotentInspect
List the voices an account can speak with: MiniMax's built-in catalogue (default, filtered by tag_list and paged) or, with is_system false, your cloned voices (all at once). Each has voice_id (for minimax_create_speech), voice_name, tag_list and a sample_audio URL; has_more and total for paging. Tags come from minimax_get_speech_config. GET /speech/voices.
Inputs: page, account, tag_list, is_system, +1 more.
minimax_run_agentRun the MiniMax agentAInspect
Give the hailuoai.video AI agent a task in plain words ("generate a 16:9 video of…", "generate an image…", "generate a song based on my lyrics") and the models it may use; it picks among them and returns the files it made. Optional reference files by public URL. Uses the account's MiniMax credits. Runs as a job: waits up to ~50 s, then returns the jobId for minimax_get_agent_job. For direct control use minimax_create_video / minimax_create_image. POST /agent.
Inputs: models (required), prompt (required), wait, account, +3 more.
minimax_stop_trackingStop tracking a jobADestructiveIdempotentInspect
Stop this API tracking a running video, image, music or speech job (an id from minimax_list_running_jobs) and free its slot on the account. It does not cancel the work at MiniMax: a video or song carries on and its get tool still finds it, but a stopped speech audioId reads as failed from then on. To ask MiniMax to stop a video use minimax_cancel_video. 404 when the id is not running. Cannot be undone. DELETE /scheduler/{id}.
Inputs: id (required).
minimax_upload_fileUpload a reference fileAInspect
Upload your own image (PNG, JPEG, WebP), audio (MP3, WAV) or MP4 video to one account, for use as a reference in minimax_create_video (fileID, end_frame_fileID, omni slots), minimax_create_image (characterFileId, referenceFileIdN) or minimax_create_template_video. Send a public https url or contentBase64 (small files), up to 50 MB here; larger files (the REST takes up to 5 GB) go to POST /files directly. Returns a fileID that works on that account only, and its ossPath. POST /files.
Inputs: url, account, mimeType, contentBase64.
minimax_upload_music_referenceUpload a reference trackAInspect
Upload an audio file (MP3 or WAV, up to 50 MB through this server) to one MiniMax account as a reference track for covers. Returns its musicId, ready at once, and the lyrics MiniMax transcribed from it (pass them as lyrics to keep the original words). A reference is a permanent record on the account, reusable for any number of covers: upload once and pass the musicId as referenceMusicId to minimax_create_music. Uploading costs no credits; a song already on the account needs no upload. POST /music/reference.
Inputs: url, name, account, mimeType, +1 more.
minimax_upload_voice_sampleUpload a voice sampleAInspect
Upload one audio sample (MP3 or WAV, 10 to 60 seconds of speech, up to 50 MB through this server) to an account for voice cloning. Returns a fileID; pass up to 10 fileIDs from the same account to minimax_clone_voice. Not for music covers: use minimax_upload_music_reference. POST /speech/upload-sample.
Inputs: url, account, mimeType, contentBase64.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
AI video, images, music & SFX: Seedance 2.5, Veo 3.1, Kling 3.0, Nano Banana Pro, 20+ models.
Kling AI: 4.0, v3 and Omni video, Image 4.0 (8K) images, avatars, lip-sync, motion control, voices
MCP server for Hailuo (MiniMax) AI video generation
AI image, video & music generation. Flux, Veo 3.1, Suno V5. Free tier included.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceOfficial MiniMax MCP server enabling AI clients to generate speech, clone voices, create videos and images, compose music, and design custom voices via MiniMax APIs.MIT
- AlicenseNot gradedqualityDmaintenanceMCP server for MiniMax's multimodal generation models, enabling text-to-speech, voice cloning, image, video, and music creation through natural language.MIT
- AlicenseAqualityNot gradedmaintenanceEnables interaction with MiniMax AI APIs for text-to-speech, voice cloning, video generation, image generation, and music creation through MCP clients like Claude Desktop and Cursor.9MIT
- AlicenseNot gradedqualityAmaintenanceEnables users to generate images and videos, synthesize speech, clone voices, and design custom voices through MiniMax APIs from MCP-compatible clients.468 npmMIT
Glama MCP Gateway
Add one secure layer between your agents and this server.