Minimax

MiniMax H3

MiniMaxToken-basedDynamic Pricing
Alias:minimax-h3
Create API Key

Compare minimax-h3 API pricing, supported endpoints, capabilities and access options on Modelsell.

视频生成文生视频多模态参考首尾帧768P2K
Starting price
View full pricing
Context
—
Maximum input window
Modalities
→

Pricing by Supplier

official
官方接口直连
768P · Input + output duration
Per second/ second
$0.08
Image In/ image
$0.04

Free reference images: 5

Other resolutions · Input + output duration
Per second/ second
$0.13
Image In/ image
$0.04

Free reference images: 5

Capabilities / Supported modalities

Vision
Input
Output

Provider & data privacy

Provider
MiniMaxDocs
Tokenizer
ABAB tokenizer
License
Proprietary (commercial)Proprietary
Data retention41 daysNot used for upstream training by default

Performance

Benchmarks

Scores on standardized evaluations. Higher percentages are better — and rank percentile shows

Metrics sourced fromArtificial Analysis 2026-09-30·MiniMax H3

This model is not included in the current benchmark snapshot.

Missing models or measurements are not zero scores.

Benchmark charts preserve the source model selection and reasoning settings. Missing models or measurements are not zero scores, and benchmark cost or speed is not this site’s service commitment.

About MiniMax H3

用文本与参考素材生成视频

MiniMax H3 使用 content 数组组织文本和参考媒体,可保留首帧、尾帧及 reference_image、reference_video、reference_audio 等角色。分辨率为 768P 或 2K,最短生成时长为 4 秒。输出为视频文件,视频参数和媒体限制由上游校验。

标准价与计量

以下为 2026-10-28 供应商目录的美元标准价;组织采购折扣不用于此处的基准计费。实际账单还会应用账户所属分组的既有倍率。成功任务按最终用量结算,失败任务按网关任务流程退款,查询不重复收费。

768P 为 $0.08/秒,2K 为 $0.13/秒。输出视频与引用视频按对应分辨率同价,结算读取 task.usage.total_seconds;参考图片前 5 张免费,超出部分 $0.04/张。

资料:供应商接口说明、供应商模型目录。具体参数由上游校验。

Use cases and prompting

原生接口

向网关 /v2/video_generation 发送以下 JSON,并使用自己的网关令牌鉴权:

{
  "model": "MiniMax-H3",
  "content": [
    {
      "type": "text",
      "text": "A red ceramic teapot rests on a wooden table in warm morning sunlight. No people, no text."
    }
  ],
  "resolution": "768P",
  "duration": 4,
  "ratio": "16:9",
  "aigc_watermark": false
}

任务为异步处理。创建响应中的 task_id 是网关公开任务 ID;使用 GET /v2/query/video_generation/{task_id} 查询,每 10–15 秒一次。成功视频见 task.content.url;文字结果见 task.content.prompt。创建、查询及源任务引用使用同一网关令牌和协议。

统一视频接口

可调用 POST /v1/videos 或 POST /v1/video/generations,使用 model: "minimax-h3"、prompt、seconds: "4"、size: "768P"。渠道将小写模型名映射为上游 MiniMax-H3。metadata.content 可指定完整媒体角色;提供后优先于自动拼接的提示词与图片。增强提示词和再生成分别选择对应模型的原生接口。

API access

API documentation

Authentication

All requests must include Authorization: Bearer <TOKEN> header. Anthropic-formatted endpoints accept the x-api-key header instead.

Generate tokens from the Tokens page; you can scope them to specific models, groups, IPs, and rate-limits.

Rate limits

SupplierRPMTPMRPD
MiniMaxUnlimitedUnlimitedUnlimited
officialUnlimitedUnlimitedUnlimited

No restriction

Frequently asked questions about minimax-h3

What is minimax-h3?

Compare minimax-h3 API pricing, supported endpoints, capabilities and access options on Modelsell.

How do I call minimax-h3?

Create an API key with access to minimax-h3, then use the exact model ID and a supported endpoint from the API access section. Request fields depend on the selected endpoint.

How is minimax-h3 priced?

Pricing depends on the selected provider group and the model billing unit. The current input, output, request, or media prices are shown on this page before sign-up.

How should I evaluate minimax-h3 for my project?

Start with the use cases and prompting guidance on this page, then evaluate the model with representative inputs from your project.