17 models · video

MiniMax API

MiniMax is a family of AI video models. varg runs 17 of them on one API key: MiniMax H3, MiniMax H3 LoRA, MiniMax H3 Max and 14 more. Between them they cover text to video, image to video, video to video and lip sync; they return video.

Price
From $0.02 per run MiniMax H3 Max Turbo Extend Video
MiniMax H3
From $0.25 per run
Flagship id
minimax_h3

Every MiniMax model runs on one varg API key and one credit balance, the same as every other model here. You pay the provider's price, with no markup.

POST /v2/tools/video/call
curl -X POST https://api.varg.ai/v2/tools/video/call \
  -H "Authorization: Bearer $VARG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "minimax_h3",
  "prompt": "A white kitten chases a butterfly across a sunlit garden. Gentle camera tracking, natural movement, soft afternoon light.",
  "aspect_ratio": "16:9",
  "duration": 5
}'
Try MiniMax H3 in the PlaygroundGet an API key

MiniMax models and prices

MiniMax on varg starts from $0.02 per run (MiniMax H3 Max Turbo Extend Video). Prices are in US dollars and are the provider's own, with no markup. A "from" price is the cheapest request the catalog states for that model; each model's page has every rate and input. The descriptions are the catalog's own, shortened.

MiniMax models on varg
ModelCatalog descriptionTakesPriceTypical time
MiniMax H3minimax_h3MiniMax H3. Frontier 2K video model with native stereo audio. Text, image, or multimodal reference inputs.Text to video, Image to videoFrom $0.25 per runfrom $0.05/s~2 min
MiniMax H3 LoRAminimax_h3_loraMiniMax H3 LoRA. H3 with your own LoRA weights (up to 3, URL or Hugging Face repo): text, image or references to video, 480p to 4K, 5-15 seconds.Text to video, Image to videoFrom $0.32 per runfrom $0.0625/s~2 min
MiniMax H3 Maxminimax_h3_maxMiniMax H3 Max. fal's post-trained H3 with stronger prompt adherence and aesthetics: text, first/last frame or multimodal references, 480p-1080p, up…Text to video, Image to videoFrom $0.05 per runfrom $0.05/s~2 min
MiniMax H3 Max 3D to Videominimax_h3_max_3d_to_videoMiniMax H3 Max 3D to Video. Turn a Blender render or 3D previs (up to 15 s) into photoreal video with the same layout, camera and timing; reference…Video to videoFrom $0.90 per runfrom $0.18/s~2 min
MiniMax H3 Max Camera Controlsminimax_h3_max_camera_controlsMiniMax H3 Max Camera Controls. Move the camera around a single image with keyframed azimuth, elevation and distance while the scene stays put. Up…Image to videoFrom $0.05 per runfrom $0.05/s~30 s
MiniMax H3 Max Extend Videominimax_h3_max_extendMiniMax H3 Max Extend Video. Continue a video with up to 15 s of prompt-guided footage; returns the full video or only the new part. 480p to 2K.Video to videoFrom $0.04 per runfrom $0.05/s~60 s
MiniMax H3 Max Insert Videominimax_h3_max_insertMiniMax H3 Max Insert Video. Generate a new 5-13 s scene between two points of an existing video from a prompt or references, then cut back to the…Video to videoFrom $0.25 per runfrom $0.05/s~2 min
MiniMax H3 Max Lip Syncminimax_h3_max_lipsyncMiniMax H3 Max Lip Sync. Talking video from one image and an audio track (5 s to 15 min); the video follows the audio length. 480p to 2K.Lip syncFrom $0.30 per runfrom $0.06/s~60 s
MiniMax H3 Max Recastminimax_h3_max_recastMiniMax H3 Max Recast. Swap the people in a 5-30 s video for people from up to four photos, keeping motion, gestures, camera, cuts and soundtrack…Video to videoFrom $1.50 per runfrom $0.30/s~3 min
MiniMax H3 Max Relightminimax_h3_max_relightMiniMax H3 Max Relight. Relight a video (up to 15 s) from a lighting-sphere image: light color, direction and softness change while subjects…Video to videoFrom $0.12 per runfrom $0.05/s~2 min
MiniMax H3 Max · 16-bit Pixelminimax_h3_max_style_16bit_pixelMiniMax H3 Max · 16-bit Pixel. 16-bit pixel-art video with audio from a prompt or a first-frame image, 768p, 5-15 seconds.Text to video, Image to videoFrom $0.40 per run$0.08/s~30 s
MiniMax H3 Max · Hand Drawnminimax_h3_max_style_hand_drawnMiniMax H3 Max · Hand Drawn. Hand-drawn pencil animation video with audio from a prompt or a first-frame image, 768p, 5-15 seconds.Text to video, Image to videoFrom $0.40 per run$0.08/s~30 s
MiniMax H3 Max · Low Polyminimax_h3_max_style_low_polyMiniMax H3 Max · Low Poly. Retro low-poly 3D video with audio from a prompt or a first-frame image, 768p, 5-15 seconds.Text to video, Image to videoFrom $0.40 per run$0.08/s~30 s
MiniMax H3 Max · Retro Toon 70sminimax_h3_max_style_retro_toon_70sMiniMax H3 Max · Retro Toon 70s. 1970s hand-painted cartoon video with audio from a prompt or a first-frame image, 768p, 5-15 seconds.Text to video, Image to videoFrom $0.40 per run$0.08/s~30 s
MiniMax H3 Max · VHSminimax_h3_max_style_vhsMiniMax H3 Max · VHS. VHS-tape video with audio from a prompt or a first-frame image, 768p, 5-15 seconds.Text to video, Image to videoFrom $0.40 per run$0.08/s~30 s
MiniMax H3 Max Turbominimax_h3_max_turboMiniMax H3 Max Turbo. Nearly H3 Max quality at half the price and a fraction of the wait, up to 1080p and 15 seconds; can sync to an audio track.Text to video, Image to videoFrom $0.03 per runfrom $0.025/s~15 s
MiniMax H3 Max Turbo Extend Videominimax_h3_max_turbo_extendMiniMax H3 Max Turbo Extend Video. Half-price Turbo version of Extend Video: up to 15 s of new footage after an existing clip, 480p to 2K.Video to videoFrom $0.02 per runfrom $0.025/s~30 s

Call the MiniMax API

The call below runs MiniMax H3. For another MiniMax model, put its id from the table in "model"; the rest of the request keeps its shape.

REST

POST https://api.varg.ai/v2/tools/video/call with Authorization: Bearer $VARG_API_KEY. It answers 202 with a job right away; poll GET /v2/jobs/{id} until its status is completed, then read output.outputs[0].url.

POST /v2/tools/video/call
curl -X POST https://api.varg.ai/v2/tools/video/call \
  -H "Authorization: Bearer $VARG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "minimax_h3",
  "prompt": "A white kitten chases a butterfly across a sunlit garden. Gentle camera tracking, natural movement, soft afternoon light.",
  "aspect_ratio": "16:9",
  "duration": 5
}'

From Claude, ChatGPT or Cursor (MCP)

The varg MCP server, https://mcp.varg.ai/mcp, runs MiniMax H3 with call_tool. Sign in with varg, no API key to paste. How to connect it

Claude Code
claude mcp add --transport http varg https://mcp.varg.ai/mcp
call_tool
{
  "tool": "video",
  "input": {
    "model": "minimax_h3",
    "prompt": "A white kitten chases a butterfly across a sunlit garden. Gentle camera tracking, natural movement, soft afternoon light.",
    "aspect_ratio": "16:9",
    "duration": 5
  },
  "wait": true
}

From a terminal (CLI)

vargai is the varg command line: every tool of the MCP server is a command. Install and log in

vargai
curl -fsSL https://varg.ai/install | sh
vargai call_tool --tool video '{"model":"minimax_h3","prompt":"A white kitten chases a butterfly across a sunlit garden. Gentle camera tracking, natural movement, soft afternoon light.","aspect_ratio":"16:9","duration":5}'

How billing works

Every model is billed at its provider's price with no markup, from one credit balance: 1 credit is $0.01. varg's fee is on the top-up, not on the run: 5.5% of the credits you buy ($0.80 at the least), or 8% for a team workspace. Pricing

MiniMax compared

Price, limits and inputs side by side, from the same catalog as this page.

How to use MiniMax

Step-by-step pages on varg how to.

MiniMax API: questions

How much does the MiniMax API cost on varg?
The lowest price the catalog states in the MiniMax family is from $0.02 per run (from $0.025/s), on MiniMax H3 Max Turbo Extend Video. Every model is billed at its provider's price with no markup, from one credit balance: 1 credit is $0.01. For the exact price of a request before you run it, send the same body to POST https://api.varg.ai/v2/estimate.
Which MiniMax models are on varg?
17: MiniMax H3 (from $0.25 per run), MiniMax H3 LoRA (from $0.32 per run), MiniMax H3 Max (from $0.05 per run), MiniMax H3 Max 3D to Video (from $0.90 per run), MiniMax H3 Max Camera Controls (from $0.05 per run), MiniMax H3 Max Extend Video (from $0.04 per run), MiniMax H3 Max Insert Video (from $0.25 per run), MiniMax H3 Max Lip Sync (from $0.30 per run), MiniMax H3 Max Recast (from $1.50 per run), MiniMax H3 Max Relight (from $0.12 per run), MiniMax H3 Max · 16-bit Pixel (from $0.40 per run), MiniMax H3 Max · Hand Drawn (from $0.40 per run), MiniMax H3 Max · Low Poly (from $0.40 per run), MiniMax H3 Max · Retro Toon 70s (from $0.40 per run), MiniMax H3 Max · VHS (from $0.40 per run), MiniMax H3 Max Turbo (from $0.03 per run) and MiniMax H3 Max Turbo Extend Video (from $0.02 per run). Each has its own page with the parameters and a call to copy.
Which MiniMax models take an image?
10 of the 17: MiniMax H3, MiniMax H3 LoRA, MiniMax H3 Max, MiniMax H3 Max Camera Controls, MiniMax H3 Max · 16-bit Pixel, MiniMax H3 Max · Hand Drawn, MiniMax H3 Max · Low Poly, MiniMax H3 Max · Retro Toon 70s, MiniMax H3 Max · VHS and MiniMax H3 Max Turbo. The catalog lists image to video for them; pass the image as a URL in "files".
What is the longest clip the MiniMax models make?
15 seconds, on MiniMax H3, MiniMax H3 LoRA, MiniMax H3 Max, MiniMax H3 Max Camera Controls, MiniMax H3 Max Extend Video, MiniMax H3 Max · 16-bit Pixel, MiniMax H3 Max · Hand Drawn, MiniMax H3 Max · Low Poly, MiniMax H3 Max · Retro Toon 70s, MiniMax H3 Max · VHS, MiniMax H3 Max Turbo and MiniMax H3 Max Turbo Extend Video, by the models' input schemas. Then MiniMax H3 Max Insert Video (13 s).
How long do MiniMax requests take?
Typically from about 15 s (MiniMax H3 Max Turbo) to about 3 min (MiniMax H3 Max Recast), by the catalog's estimates for each model's fastest route. The API answers at once with a job id, so nothing waits on an open connection while it runs.
How do I call the MiniMax API?
Send POST https://api.varg.ai/v2/tools/video/call with the header "Authorization: Bearer $VARG_API_KEY" and a JSON body that includes "model", for example "model": "minimax_h3". The call answers 202 with a job right away. Poll GET https://api.varg.ai/v2/jobs/{id} until "status" is "completed", then read output.outputs[0].url, or pass options.webhook_url to be called back.
Can I use MiniMax from Claude, ChatGPT or Cursor?
Yes. Add the varg MCP server (https://mcp.varg.ai/mcp) to Claude, ChatGPT, Cursor or any other MCP client and sign in with varg, with no API key to paste. The agent runs a model with call_tool, passing tool "video" and the model id, for example "model": "minimax_h3".

More model APIs

Same API key, same request shape: change the model string to switch.

All models

For AI agents: each model above has its full spec as Markdown at varg.ai/models/minimax_h3.md (swap the id), and every model is listed in varg.ai/models/llms.txt.