# Video generation: 30-second clips with native audio change the planning math

- **Published:** 2026-07-10
- **Updated:** 2026-10-12
- **Category:** Models & Benchmarks
- **Author:** Sinthetix
- **URL:** https://sinthetix.com/blog/video-generation-longer-clips-native-audio

> ByteDance says Seedance 2.5 makes 30-second clips with joint audio; Google's Veo 3.1 page cites 8-second clips. All claims are vendor-stated.

The video-generation story in mid-2026 is less about image quality and more about length, sound and control. Two vendor pages show the contrast.

**Key takeaways**

- ByteDance describes Seedance 2.5 as generating up to 30 seconds in one pass, extendable twice, with joint audio-video generation [1].
- Google's Veo 3.1 page says "Veo videos are 8 seconds long" (a benchmark footnote), with native audio, 1080p and 4K output, SynthID watermarking and scene extension to chain clips [2].
- Most specifications for Seedance 2.5 are vendor claims; a secondary roundup notes they are not independently verified [3].

## What each vendor says

ByteDance's Seedance 2.5 page calls it "built for 30-second storytelling." It generates up to 30 seconds in a single generation, can be extended twice, and is described as an audio-video joint generation model. Editing features include green-screen editing, camera movement and performance blocking, and it reads reference videos for framing and cinematic language [1]. The page does not state resolution, pricing or rate limits. The ByteDance page shows no release date. A June 26 roundup (secondary source) reports the model was announced on June 23, 2026 at ByteDance's Volcano Engine FORCE conference, in enterprise beta with public launch targeted for early July, and that the 30-second figures are vendor-stated and not yet independently verified [3].

Google's Veo 3.1 page says the model generates audio natively, including sound effects, ambient noise and dialogue, outputs at 1080p and 4K, and marks every video with SynthID. Its footnote to a human-rater comparison states that "Veo videos are 8 seconds long," and the page says clips can be extended into longer videos using the last second of a shot [2]. The page's performance charts are labelled "Last updated October 2025."

|  | Seedance 2.5 (ByteDance) | Veo 3.1 (Google) |
| --- | --- | --- |
| Single-pass length | Up to 30 seconds [1] | 8 seconds [2] |
| Longer output | Extend twice [1] | Scene extension [2] |
| Audio | Joint audio-video generation [1] | Native sound effects, ambient noise, dialogue [2] |
| Resolution | Not stated [1] | 1080p and 4K [2] |
| Provenance marking | Not stated [1] | SynthID watermark [2] |

## Why length and audio matter to buyers

A longer single pass means fewer stitched shots, which is where continuity errors tend to appear. Native audio removes a separate voiceover and sound-design step for drafts. Neither makes the output finished work: Google's page says natural and consistent spoken audio, particularly for shorter speech segments, remains an area of active development [2].

Provenance is the quieter issue. Veo documents a watermark; the Seedance page we reviewed does not mention one. If generated video will carry your brand, ask any vendor how outputs are marked and what usage rights apply.

## Practical implication

Treat 30-second generation as a promising pilot for storyboards and internal training clips, not a production commitment, until the claims are independently tested on your own content. Check for provenance marking and review every clip before anything customer-facing goes out.

## Sources

1. [Seedance 2.5](https://seed.bytedance.com/en/seedance2_5), ByteDance Seed
2. [Veo](https://deepmind.google/models/veo/), Google DeepMind
3. [Seedance 2.5 vs Veo 3.1 vs Kling 3.0: Best AI Video (July 2026)](https://blog.buildfastwithai.com/seedance-2-5-vs-veo-3-1-vs-kling-3-0-best-ai-video-2026), Build Fast with AI, 2026-06-26

## How Sinthetix helps

**Generative AI & LLM**: Integrate powerful AI models to turn unstructured data into insights and power chatbots.

- RAG systems
- LLM-powered chatbots
- Document analysis

[Explore Generative AI & LLM](https://sinthetix.com/services/generative-ai-llm) | [Talk to us](https://sinthetix.com/contact)

Sinthetix is an AI engineering studio. We ship agents, automation, and internal tools for operators who need results, not reports. [About Sinthetix](https://sinthetix.com/about)