Skip to content

MiniMax ​

Modes: video · audio · text · Models: 19

Vendor: MiniMax · Official API docs: MiniMax Platform — Video Generation

MiniMax provides MiniMax H3 (formerly Hailuo 03) for text-to-video and image-to-video at up to 2K, with start/end frames plus image, video, and audio references. The catalog also includes MiniMax H3 Max, the Hailuo 2.3 family, and MiniMax Music v2 for prompt-driven music with vocals.

Models ​

idNameInput type
hailuo-2.3Hailuo 2.3t2v
hailuo-2.3-proHailuo 2.3 Prot2v
hailuo-2.3-fastHailuo 2.3 Fasti2v
hailuo-2.3-fast-proHailuo 2.3 Fast Proi2v
minimax-h3MiniMax H3t2v
minimax-h3-highspeedMiniMax H3 Highspeedt2v
minimax-h3-liteMiniMax H3 Litet2v
minimax-music-v2MiniMax Music v2music
minimax-music-v3MiniMax Music v3music
minimax-h3-maxMiniMax H3 Maxt2v
minimax-h3-max-turboMiniMax H3 Max Turbot2v
minimax-h3-max-camera-controlsMiniMax H3 Max Camera Controlsi2v
minimax-h3-max-lip-syncMiniMax H3 Max Lip Synci2v
minimax-h3-max-extendMiniMax H3 Max Extendv2v
minimax-h3-max-relightMiniMax H3 Max Relightv2v
minimax-h3-max-recastMiniMax H3 Max Recastv2v
minimax-m3MiniMax M3i2t
minimax-m3.1-flashMiniMax M3.1 Flashi2t
minimax-h3-context-irMiniMax H3 Context IRv2t

CLI ​

bash
# text-to-video with a camera directive
gen-ai generate -m hailuo-2.3 \
  -p "a cat walks forward through a neon alley at night [Push in]" \
  -d 10

# image-to-video from a start frame (Fast variant)
gen-ai generate -m hailuo-2.3-fast -p "she turns and smiles" -i ./frame.png -d 6

# MiniMax H3 with a start frame
gen-ai generate -m minimax-h3 -p "the camera circles the subject" \
  --start-frame ./frame.png -d 10

# music with vocals
gen-ai generate -m minimax-music-v2 \
  -p "upbeat indie pop with bright guitars and an anthemic chorus, female vocals"

MCP ​

json
{ "name": "picsart_generate",
  "arguments": {
    "model": "hailuo-2.3",
    "prompt": "a cat walks forward through a neon alley at night [Push in]",
    "duration": 10
  } }
json
{ "name": "picsart_generate",
  "arguments": {
    "model": "minimax-music-v2",
    "prompt": "upbeat indie pop with bright guitars and an anthemic chorus, female vocals"
  } }

Parameters ​

Full parameter surface for every model, sourced from gen-ai models info <id> --json. CLI flags show the primary short form; the canonical --kebab-case long form always works too.

hailuo-2.3 — Hailuo 2.3 ​

Try hailuo-2.3 in Playground ↗

Input type: t2v

ParamCLI flagTypeValues
prompt-ptextrequired
enhancePrompt--enhance-promptbooleantrue · false (default true)
duration-denum6 · 10 (default 6)
imageUrls-ifileimage (up to 1)

hailuo-2.3-pro — Hailuo 2.3 Pro ​

Try hailuo-2.3-pro in Playground ↗

Input type: t2v

ParamCLI flagTypeValues
prompt-ptextrequired
enhancePrompt--enhance-promptbooleantrue · false (default true)
imageUrls-ifileimage (up to 1)

hailuo-2.3-fast — Hailuo 2.3 Fast ​

Try hailuo-2.3-fast in Playground ↗

Input type: i2v

ParamCLI flagTypeValues
prompt-ptextrequired
enhancePrompt--enhance-promptbooleantrue · false (default true)
duration-denum6 · 10 (default 6)
imageUrls-ifilerequired image (up to 1)

hailuo-2.3-fast-pro — Hailuo 2.3 Fast Pro ​

Try hailuo-2.3-fast-pro in Playground ↗

Input type: i2v

ParamCLI flagTypeValues
prompt-ptextrequired
enhancePrompt--enhance-promptbooleantrue · false (default true)
imageUrls-ifilerequired image (up to 1)

minimax-h3 — MiniMax H3 ​

Try minimax-h3 in Playground ↗

Input type: t2v

ParamCLI flagTypeValues
prompt-ptextrequired (≤7000 chars)
startFrame--start-framefileimage
endFrame--end-framefileimage
imageUrls-ifileimage (up to 9)
videoUrls--video-urlsfilevideo (up to 3)
audioUrls--audio-urlsfileaudio (up to 3)
resolution-renum768P · 2K (default 2K)
duration-drange5–15, step 1 (default 5)
aspectRatio--arenumadaptive · 21:9 · 16:9 · 4:3 · 1:1 · 3:4 · 9:16 (default adaptive)

minimax-h3-highspeed — MiniMax H3 Highspeed ​

Try minimax-h3-highspeed in Playground ↗

Input type: t2v

ParamCLI flagTypeValues
prompt-ptextrequired (≤7000 chars)
startFrame--start-framefileimage
endFrame--end-framefileimage
imageUrls-ifileimage (up to 9)
videoUrls--video-urlsfilevideo (up to 3)
audioUrls--audio-urlsfileaudio (up to 3)
resolution-renum480P · 768P (default 768P)
duration-drange5–15, step 1 (default 5)
aspectRatio--arenumadaptive · 21:9 · 16:9 · 4:3 · 1:1 · 3:4 · 9:16 (default adaptive)

minimax-h3-lite — MiniMax H3 Lite ​

Try minimax-h3-lite in Playground ↗

Input type: t2v

ParamCLI flagTypeValues
prompt-ptextrequired (≤7000 chars)
startFrame--start-framefileimage
endFrame--end-framefileimage
imageUrls-ifileimage (up to 9)
videoUrls--video-urlsfilevideo (up to 3)
audioUrls--audio-urlsfileaudio (up to 3)
resolution-renum480P · 768P (default 768P)
duration-drange5–15, step 1 (default 5)
aspectRatio--arenumadaptive · 21:9 · 16:9 · 4:3 · 1:1 · 3:4 · 9:16 (default 16:9)

minimax-music-v2 — MiniMax Music v2 ​

Try minimax-music-v2 in Playground ↗

Input type: music

ParamCLI flagTypeValues
prompt-ptextrequired (≤2000 chars)
lyricsPrompt--lyrics-prompttextlyrics, or a description of the lyrical theme (≤2000 chars, min 10)
lyricsOptimizer--lyrics-optimizerbooleantrue · false (default false)
isInstrumental--is-instrumentalbooleantrue · false (default false)
outputFormat--output-formatenumurl · hex (default url)

Notes: Pro variants omit duration (fixed 6s at 1080p); 10s is available at 768p only. On the Fast image-to-video models imageUrls is required.

minimax-music-v3 — MiniMax Music v3 ​

Try minimax-music-v3 in Playground ↗

Input type: music

ParamCLI flagTypeValues
prompt-ptextrequired (≤2000 chars)
lyricsPrompt--lyrics-prompttextfree text (≤2000 chars)
lyricsOptimizer--lyrics-optimizerbooleantrue · false (default false)
isInstrumental--is-instrumentalbooleantrue · false (default false)
sampleRate--sample-rateenum16000 · 24000 · 32000 · 44100 (default 44100)
bitrate--bitrateenum32000 · 64000 · 128000 · 256000 (default 256000)
format--formatenummp3 · wav · pcm (default mp3)

minimax-h3-max — MiniMax H3 Max ​

Try minimax-h3-max in Playground ↗

Input type: t2v

ParamCLI flagTypeValues
prompt-ptextrequired (≤50000 chars)
startFrame--start-framefileimage
endFrame--end-framefileimage
imageUrls-ifileimage (up to 9)
videoUrls--video-urlsfilevideo (up to 3)
audioUrls--audio-urlsfileaudio (up to 3)
resolution-renum480p · 768p · 1080p (default 768p)
duration-drange5–15, step 1 (default 5)
aspectRatio--arenumadaptive · 21:9 · 16:9 · 4:3 · 1:1 · 3:4 · 9:16 (default 16:9)
promptExpansionMode--prompt-expansion-modeenumbalanced · quality (default balanced)
seed--seedrange-1–2147483647 (default -1)
enableSafetyChecker--enable-safety-checkerbooleantrue · false (default true)

minimax-h3-max-turbo — MiniMax H3 Max Turbo ​

Try minimax-h3-max-turbo in Playground ↗

Input type: t2v

ParamCLI flagTypeValues
prompt-ptextrequired (≤50000 chars)
startFrame--start-framefileimage
endFrame--end-framefileimage
resolution-renum480p · 768p · 1080p (default 768p)
duration-drange5–15, step 1 (default 5)
aspectRatio--arenum21:9 · 16:9 · 4:3 · 1:1 · 3:4 · 9:16 (default 16:9)
promptExpansionMode--prompt-expansion-modeenumbalanced · quality (default balanced)
seed--seedrange-1–2147483647 (default -1)
enableSafetyChecker--enable-safety-checkerbooleantrue · false (default true)

minimax-h3-max-camera-controls — MiniMax H3 Max Camera Controls ​

Try minimax-h3-max-camera-controls in Playground ↗

Input type: i2v

ParamCLI flagTypeValues
prompt-ptextfree text (≤50000 chars)
startFrame--start-framefilerequired image
resolution-renum480p · 768p · 1080p (default 480p)
duration-drange5–15, step 1 (default 5)
cameraTrajectory--camera-trajectory-time · --camera-trajectory-azimuth · --camera-trajectory-elevation · --camera-trajectory-distanceobject[]{time, azimuth, elevation, distance}
promptExpansionMode--prompt-expansion-modeenumbalanced · quality (default balanced)
seed--seedrange-1–2147483647 (default -1)
enableSafetyChecker--enable-safety-checkerbooleantrue · false (default true)

minimax-h3-max-lip-sync — MiniMax H3 Max Lip Sync ​

Try minimax-h3-max-lip-sync in Playground ↗

Input type: i2v

ParamCLI flagTypeValues
startFrame--start-framefilerequired image
audioUrl-afilerequired audio
resolution-renum480p · 768p · 1080p · 2k (default 768p)
enableTranscription--enable-transcriptionbooleantrue · false (default false)
seed--seedrange-1–2147483647 (default -1)
enableSafetyChecker--enable-safety-checkerbooleantrue · false (default true)

minimax-h3-max-extend — MiniMax H3 Max Extend ​

Try minimax-h3-max-extend in Playground ↗

Input type: v2v

ParamCLI flagTypeValues
prompt-ptextrequired
videoUrl--videofilerequired video
duration-drange5–15, step 1 (default 5)
output--outputenumextended (Source + continuation) · continuation (Continuation only) (default extended)
aspectRatio--arenumauto · 21:9 · 16:9 · 4:3 · 1:1 · 3:4 · 9:16 (default auto)
resolution-renum480p · 768p · 1080p · 2k (default 768p)
enhancePrompt--enhance-promptbooleantrue · false (default true)
enableSafetyChecker--enable-safety-checkerbooleantrue · false (default true)
seed--seedrange0–2147483647, step 1

minimax-h3-max-relight — MiniMax H3 Max Relight ​

Try minimax-h3-max-relight in Playground ↗

Input type: v2v

ParamCLI flagTypeValues
videoUrl--videofilerequired video
imageUrls-ifilerequired image (up to 1)
resolution-renum480p · 768p · 1080p · 2k (default 768p)
aspectRatio--arenumadaptive · 21:9 · 16:9 · 4:3 · 1:1 · 3:4 · 9:16 (default 16:9)
seed--seedrange0–2147483647, step 1
enableSafetyChecker--enable-safety-checkerbooleantrue · false (default true)

minimax-h3-max-recast — MiniMax H3 Max Recast ​

Try minimax-h3-max-recast in Playground ↗

Input type: v2v

ParamCLI flagTypeValues
prompt-ptextfree text (≤2000 chars)
videoUrl--videofilerequired video
imageUrls-ifilerequired image (up to 4)
resolution-renum768p · 1080p (default 1080p)
enhanceRealism--enhance-realismbooleantrue · false (default false)
seed--seedrange0–2147483647, step 1
enableSafetyChecker--enable-safety-checkerbooleantrue · false (default true)

minimax-m3 — MiniMax M3 ​

Try minimax-m3 in Playground ↗

Input type: i2t

ParamCLI flagTypeValues
prompt-ptextrequired
imageUrls-ifileimage (up to 8)

minimax-m3.1-flash — MiniMax M3.1 Flash ​

Try minimax-m3.1-flash in Playground ↗

Input type: i2t

ParamCLI flagTypeValues
prompt-ptextrequired
imageUrls-ifileimage (up to 8)
thinking--thinkingenumlow · medium · high (default high)

minimax-h3-context-ir — MiniMax H3 Context IR ​

Try minimax-h3-context-ir in Playground ↗

Input type: v2t

ParamCLI flagTypeValues
prompt-ptextrequired
startFrame--start-framefileimage
endFrame--end-framefileimage
imageUrls-ifileimage (up to 9)
videoUrls--video-urlsfilevideo (up to 3)
audioUrls--audio-urlsfileaudio (up to 3)
duration-drange4–15, step 1 (default 5)
aspectRatio--arenumadaptive · 21:9 · 16:9 · 4:3 · 1:1 · 3:4 · 9:16 (default adaptive)

Pricing ​

bash
gen-ai pricing hailuo-2.3 --duration 10

Cost scales with duration and resolution (the Pro variants run at 1080p; 10s is 768p only). Music is priced per generated track.

Built on @picsart/ai-sdk · gen-ai CLI · Picsart MCP · Media Studio · Skills