Skip to content

MiniMaxVideo modelChecking

MiniMax H3 API

Model IDminimax-h3-image-to-video

MiniMax H3 video generation with text, image, multimodal reference and regeneration variants.

  • Image to video
  • Text to video
  • Video to video
  • Reference to video
  • Up to 9 input images
  • Up to 15 seconds
  • 480p or 768p output

Run MiniMax H3 in the playground

The form is the request body: every field shows its key. Sign in to run it; a run is billed like an API task.

Parameters

Model
model

MiniMax H3Image to videominimax-h3-image-to-video

Required
input.prompt

Text prompt for video generation. ≤ 20480 chars

input.resolution

Output resolution.

input.duration

Duration in seconds, 5-15.

Common options
input.firstFrameUrl

first frame image HTTPS image URL or media:<uuid>

input.lastFrameUrl

last frame image HTTPS image URL or media:<uuid>

Advanced (1)
input.aigc_watermark

Whether to add an AIGC watermark to the generated video.

Estimated cost
Loading price…

Credits are reserved when a task is accepted and charged only on success. Failed tasks release the reservation; unconfirmed outcomes remain reserved until verified.

Request preview

{"model": "minimax-h3-image-to-video","input": {"resolution": "2K","duration": "5","aigc_watermark": false}}

The web playground sends this body with your signed-in session. From code, send the same body with an API key.

Result

Run the request and the result appears here, with its status and cost.

MiniMax H3 pricing

Prices are shown in USD for each available variant and specification. The billing unit is shown with its rate; an official reference appears only when a matching list price is published.

Loading price…
  • Charged only on successOnly successful tasks are charged. Failed tasks release reserved credits; tasks with an unconfirmed outcome keep their reservation until the result is verified.
  • Safe retriesIdempotent retries with the same key do not create a second charge.
  • Price changesSend expectedPriceMicros to reject the task with PRICE_CHANGED if the price changed after you displayed it.

Compare MiniMax H3 variants

Each variant has its own model ID and input schema. Pick the task first, then the tier. Additional fields are shown when variants differ.

  • Max turbo image to video

    minimax-h3-max-turbo-image-to-video

    Image to video; Text, image; 5–15 s; 480p, 768p

  • Max turbo text to videoDefault

    minimax-h3-max-turbo-text-to-video

    Text to video; Text; 5–15 s; 480p, 768p; 21:9, 16:9, 4:3, 1:1, 3:4, 9:16

    aspectRatio
  • Max image to video

    minimax-h3-max-image-to-video

    Image to video; Text, image; 5–15 s; 480p, 768p

  • Max text to video

    minimax-h3-max-text-to-video

    Text to video; Text; 5–15 s; 480p, 768p; 21:9, 16:9, 4:3, 1:1, 3:4, 9:16

    ratio
  • Image to video

    minimax-h3-image-to-video

    Image to video; Text, image; 5–15 s; 2K, 768p

    aigc_watermark
  • Text to video

    minimax-h3-text-to-video

    Text to video; Text; 5–15 s; 2K, 768p; 21:9, 16:9, 4:3, 1:1, 3:4, 9:16

    ratioaigc_watermark
  • Regeneration text to video

    minimax-h3-regeneration-text-to-video

    Video to video; Text, video; 2K

  • Regeneration image to video

    minimax-h3-regeneration-image-to-video

    Video to video; Text, image, video; 2K

    firstFrameUrllastFrameUrl
  • Multimodal reference to video

    minimax-h3-multimodal-reference-to-video

    Reference to video; Text, image, video, audio; 5–15 s; 2K, 768p; 21:9, 16:9, 4:3, 1:1, 3:4, 9:16

What MiniMax H3 can do

MiniMax describes H3 as a multimodal video model that can use text, images, video and audio and generate video with native stereo sound. MiniMax's model documentation describes up to 2K and 15 seconds; actual options depend on the Hermes variant.

The Hermes catalog separates H3, Max and Max Turbo text-to-video and image-to-video, multimodal reference-to-video, and regeneration tasks. The multimodal reference schema lists up to nine images, three videos and three audio inputs. Regeneration variants take a base video; the other variants use different fields. Check the selected variant's status, valid options and price before running.

Capabilities

  • Up to 15 seconds

    Choose a length from 5 to 15 seconds.

    duration

    Every variant except Regeneration text to video, Regeneration image to video

  • Start and end frames

    Set the first frame and, optionally, the last frame the clip moves to.

    firstFrameUrllastFrameUrl

    Variants: Max turbo image to video, Max image to video, Image to video, Regeneration image to video

  • 480p or 768p output

    Pick the output resolution: 480p or 768p.

    resolution

    Every variant except Regeneration text to video, Regeneration image to video

  • 6 aspect ratios

    21:9, 16:9, 4:3, 1:1, 3:4 or 9:16.

    aspectRatioratio

    Variants: Max turbo text to video, Max text to video, Text to video, Multimodal reference to video

  • Up to 9 input images

    Send up to 9 images in one request as HTTPS URLs or uploaded media references.

    imageUrls

    Variants: Multimodal reference to video

  • Video input

    Accepts a source video as input.

    baseVideoUrlvideoUrls

    Variants: Regeneration text to video, Regeneration image to video, Multimodal reference to video

  • Audio input

    Accepts reference audio as input.

    audioUrls

    Variants: Multimodal reference to video

Use cases

  1. Generate a scene

    Describe the visual action and sound for a text-to-video task.

  2. Create from references

    Use the multimodal variant to supply supported reference media.

  3. Regenerate a clip

    Select a regeneration variant when you have a base video and follow its input requirements.

Call MiniMax H3 from code

Every image, video and music model uses the same asynchronous job API: submit a task, then poll for the result.

  1. 1

    Create an API key

    In the dashboard, with the jobs:create and jobs:read scopes. API keys
  2. 2

    Submit a task

    POST /api/v1/jobs with the model ID and an input object. The response is 202 with a task ID.
  3. 3

    Poll for the result

    GET /api/v1/jobs/{id} until the status is succeeded, then download the output URLs.

Minimal request

Required fields only, generated from the schema of Image to video. Set HERMES_API_KEY in your environment.

Model ID
minimax-h3-image-to-video
Endpoint
https://api.hermes-ai.net/api/v1/jobs
Full task API reference
# 1. Submit the task (asynchronous; returns 202 with a task id)
curl -X POST https://api.hermes-ai.net/api/v1/jobs \
  -H "Authorization: Bearer $HERMES_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{
  "model": "minimax-h3-image-to-video",
  "input": {
    "prompt": "A lighthouse on a rocky coast at sunrise, soft light",
    "resolution": "2K",
    "duration": "5"
  }
}'

# 2. Poll until status is succeeded, failed or reconciliation_required
curl https://api.hermes-ai.net/api/v1/jobs/TASK_ID \
  -H "Authorization: Bearer $HERMES_API_KEY"

Input parameters

Fields of Image to video (minimax-h3-image-to-video). Unknown fields are rejected with INVALID_MODEL_INPUT.

FieldTypeRequiredDefaultAllowed values and constraints
prompt
stringrequired—

Text prompt for video generation.

≤ 20480 chars

firstFrameUrl
stringoptional—

first frame image

HTTPS image URL or media:<uuid>

lastFrameUrl
stringoptional—

last frame image

HTTPS image URL or media:<uuid>

resolution
stringrequired"2K"

Output resolution.

One of: "2K", "768P"

duration
stringrequired"5"

Duration in seconds, 5-15.

One of: "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15"

aigc_watermark
booleanoptionalfalse

Whether to add an AIGC watermark to the generated video.

Upload a local file first

Replace the example URL with the returned media reference.

POST /api/v1/media
curl -X POST https://api.hermes-ai.net/api/v1/media \
  -H "Authorization: Bearer $HERMES_API_KEY" \
  -H "Content-Type: image/png" \
  --data-binary @input.png
# → { "reference": "media:<uuid>", ... }  Use the reference in an image, video or audio field.
Headers, response, task statuses and errors
HeaderValueNotes
AuthorizationBearer <API key>Or send the key in x-api-key.
Content-Typeapplication/json
Idempotency-Key<unique string ≤128>Required. Reuse it only to retry the identical request.
FieldTypeNotes
idstringHermes AI task ID. Use it to poll.
modelstringModel slug.
statusstringSee task statuses.
errorCodestring | nullSet when a task fails or needs review.
priceMicrosintegerQuote fixed at admission, USD micro-units.
currency"USD"
createdAt / updatedAtstringISO 8601 timestamps.
assets[]array{ id, type, url?, text? }. url is a same-origin /api/v1/results/<uuid> link.
replayedbooleanSubmit only. true when an idempotent retry matched.
StatusTerminalMeaning
queuednoAccepted and waiting for a worker.
submittingnoBeing sent to the model provider.
runningnoThe model is generating.
succeededyesFinished. Results are in assets.
failedyesFinished without a result. The hold is released.
reconciliation_requiredyesThe result is being verified. Keep the task ID and do not resubmit automatically.
HTTPCodeMeaning and action
401AUTH_REQUIREDMissing or invalid API key or session.
403INSUFFICIENT_SCOPEThe API key lacks the scope this endpoint needs.
403PERMISSION_DENIEDThis credential cannot perform the operation.
429API_KEY_RATE_LIMITEDThe API key's rate limit was reached. Honor Retry-After.
429API_KEY_USAGE_EXCEEDEDThe API key reached its usage limit. Create a new key.
400INVALID_REQUESTThe body is not a JSON object or a required value is missing.
400INVALID_IDEMPOTENCY_KEYIdempotency-Key must be 1–128 printable ASCII characters.
400INVALID_MODEL_INPUTinput does not match the model schema (unknown field, missing required field, or invalid value).
400 / 413INPUT_TOO_LARGEThe request body is larger than 256 KB. Upload media first and send a media reference.
404REQUEST_NOT_FOUNDNo chat request with this ID belongs to your account.
404MODEL_NOT_FOUNDNo model has this slug.
404JOB_NOT_FOUNDNo task with this ID belongs to your account.
409IDEMPOTENCY_CONFLICTThis Idempotency-Key was used with different input. Use a new key for a new request.
409PRICE_CHANGEDThe price differs from expectedPriceMicros. Fetch the price and confirm again.
409MODEL_PRICE_UNAVAILABLENo price is configured for this model yet, so it cannot run.
409PAYMENT_REVIEW_REQUIREDA refund left an uncovered balance. Contact support or cover the balance before submitting tasks.
409INSUFFICIENT_CREDITSAvailable credits do not cover the task price.
429RATE_LIMITEDToo many task submissions per minute. Retry later.
429EXPORT_RATE_LIMITEDToo many CSV exports this minute. Honor Retry-After.
429CONCURRENCY_LIMITEDToo many unfinished tasks. Wait for one to finish.
503EXECUTION_UNAVAILABLEExecution is not configured for this model right now.
429UPSTREAM_RATE_LIMITEDThe model provider is rate limiting chat requests. Retry with backoff; nothing was charged.
502UPSTREAM_ERRORThe model provider failed the chat request. Nothing was charged; retry later.
503PLATFORM_BUDGET_EXCEEDEDThis model is temporarily unavailable. Try again later.
503EXECUTION_TEMPORARILY_UNAVAILABLEExecution is temporarily unavailable. Retry with the same Idempotency-Key.
503AUTH_UNAVAILABLEAuthentication is temporarily unavailable. Retry later.
503API_UNAVAILABLEThe API is temporarily unavailable. Retry later.

MiniMax H3 API FAQ

Answers built from the model's schema and the Hermes AI API contract.

Can all H3 variants accept audio references?

No. Audio reference fields are listed for the multimodal reference-to-video variant; the other tasks have different input schemas.

What can the MiniMax H3 API do?

MiniMax H3 by MiniMax handles image to video, text to video, video to video and reference to video. The catalog lists 9 variants, each with its own model ID and input schema. Check the current availability before sending a request.

Which durations, resolutions and aspect ratios does MiniMax H3 support?

Across its variants, MiniMax H3 supports durations from 5 to 15 seconds, 480p, 768p and 2K resolution and aspect ratios 21:9, 16:9, 4:3, 1:1, 3:4 and 9:16. Each variant accepts its own subset; the variants table on this page lists them.

What input does MiniMax H3 accept?

A text prompt of up to 40,000 characters, plus image, video or audio input (up to 9 images per request). Media fields take a public HTTPS URL or a media:<uuid> reference returned by POST /api/v1/media, so local files can be uploaded first. Unknown fields are rejected with INVALID_MODEL_INPUT.

How is MiniMax H3 priced and billed?

You pay per successful task from prepaid credits, with no subscription. The price depends on the variant and the specification you choose, such as resolution or duration; see the pricing table on this page or the pricing page. Credits are reserved when a task is accepted and charged only when it succeeds; a failed task returns the reservation to your balance.

Which MiniMax H3 variant should I choose?

For image to video, Max turbo image to video, Max image to video and Image to video differ in fields: Image to video adds aigc_watermark; for text to video, Max turbo text to video, Max text to video and Text to video differ in fields: Max turbo text to video adds aspectRatio; Max text to video adds ratio; Text to video adds ratio, aigc_watermark; for video to video, Regeneration text to video and Regeneration image to video differ in fields: Regeneration image to video adds firstFrameUrl, lastFrameUrl; for reference to video, use Multimodal reference to video. Pick the task first, then compare the tiers in the variants table.

How do I call the MiniMax H3 API from code?

Send POST https://api.hermes-ai.net/api/v1/jobs with the variant's model ID and an input object, then poll GET /api/v1/jobs/{id} until the status is succeeded or failed. The API section has cURL, JavaScript and Python examples generated from the schema.

Can I try MiniMax H3 without writing code?

Yes. Sign in and use the playground on this page: it sends the same request as the API with your session and charges the same price. Code needs an API key from the dashboard.

What happens when a task fails or I hit a rate limit?

A failed task is not charged and its reserved credits are returned. Too many submissions per minute return 429 RATE_LIMITED, and too many unfinished tasks return 429 CONCURRENCY_LIMITED; retry later or wait for a task to finish.