Sume video model ids in one table (October 2026)

All Sume video ids with resolution and length limits: Seedance, Kling 3, Wan 3.0, H3, H3 Max, Omni, Grok, plus Google Cloud and Tencent rows. Read 2026-10-13.

6 min readSume
All posts

Sume's Video Router doc catalog in code has 18 generate ids (10 standard, 8 conditional) plus the edit rows below, from seedance-2.5 to hunyuan-1.5; some of them appear only where a provider is configured. Use this table as a map, then confirm with GET /v1/videos/models. Kling 4.0 and Vidu Q4 are not in it.

Always-listed rows

These are the standard rows.

Standard rows from the Sume Video Router docs, read 2026-10-13.
IdResolutionsLength
seedance-2.5480p, 720p, 1080p4-30 s
seedance-2480p, 720p, 1080p4-15 s
seedance-2-fast480p, 720p4-15 s
seedance-2-mini480p, 720p4-15 s
kling-31080p4-15 s
wan-3.0480p, 720p, 1080p2-30 s
grok-imagine-video-1.5480p, 720p4-15 s
minimax-h3480p, 768p, 2K, 4K5-15 s
minimax-h3-max480p, 768p, 1080p5-15 s
gemini-omni-flash-1.1360p, 720p, 1080p, 4K3-10 s

Conditional rows

These need a provider that your environment may not have.

Conditional rows from the Sume Video Router docs, read 2026-10-13.
IdNeedsLimits
veo-3.1, veo-3.1-fast, veo-3.1-liteGoogle CloudText-to-video, 720p, 4/6/8 s
wand-vega-1.0-liteTencent VODText-to-video, 768p-4K, 4-15 s
wand-vega-1.0-pro, wand-vega-1.5-proTencent VODText-to-video, 720p-4K, 4-15 s
mingmou-1Tencent VODText-to-video, 16:9, 5 or 10 s, silent
hunyuan-1.5Tencent VODText-to-video, 5 s, silent

Edit and other rows

higgsfield-genjutsu, h3-max-recast and hailuo-h3-regen take a source clip and are never picked by sume/auto; see the edit post.

Sources

Related posts

More in Models

All Models posts

Written by Sume