Skip to content

AI model

Seedance

ByteDance's family of AI video-generation models, which turns text and image prompts into short video clips; it competes with Kling, Runway and Sora.

Known aliases

  • ByteDance Seedance
  • Seedance 2.0
  • Seedance 2.5

Current stories

investOne report1 publisher

Just 2,886 of Douyin's 221,900 new AI dramas reached the break-even line

Just 1.3% of the 221,900 AI dramas uploaded to Douyin in the first half of 2026 cleared the 50 million-view break-even line, according to an industry report. Viewing is concentrating in fewer hits while cities subsidise the compute behind every title.

Reality

Evidence35
Adoption72
Hype gap+15
Incentives70
Confidence38
buildOne report1 publisher

Making three sign-in paths resolve to one token wallet cost a month of Dantiva's build

Dantiva's video generation is three Google model calls. The schedule went to accounts, a reserve-capture-refund token ledger, support tooling, and making one wallet work behind three different sign-ins.

Publishers:dev.to

Reality

Evidence42
Adoption14
Hype gap+8
Incentives64
Confidence46
buildOne report1 publisher

Pooling per-job status polls into one Server Action cuts the client queue depth to one

Next.js dispatches Server Actions one at a time per client connection, so open-higgsfield collapses every in-flight generation poll into a single 4,000 ms tick and fans the status calls out on the server with Promise.all.

Publishers:dev.to

Reality

Evidence45
Adoption15
Hype gap+35
Incentives60
Confidence55
investOne report1 publisher

Runway's ARR doubled to $200M in five months, driven by enterprise growth

The revenue Runway names is advertising and marketing production, and the Paris robotics team it just absorbed has produced none of it, which makes world models an option that ad money is currently funding.

Publishers:techtimes.com

Reality

Evidence33
Adoption62
Hype gap+41
Incentives74
Confidence47
productOne report1 publisher

ByteDance's reported spatial video model moves the XR rendering bill into the data centre

Bloomberg says the model would stream worlds at about 20 frames a second with 50 milliseconds of latency. That is one frame of headroom, and every minute a Pico user spends inside one is compute somebody has to pay for.

Publishers:thenextweb.com

Reality

Evidence30
Adoption18
Hype gap+22
Incentives60
Confidence48
investOne report1 publisher

Open weights take 29% of gateway tokens on a twenty-fifth of the dollars

Vercel's June data shows enterprise AI volume and spend pulling apart, with cheap open models absorbing routine work while Anthropic holds 61% of the money and 72% or more of the jobs that hurt when they go wrong.

Publishers:vercel.com

Reality

Evidence55
Adoption68
Hype gap+22
Incentives78
Confidence52
buildOne report1 publisher

ByteDance's MPA deal buys peace at the output layer and leaves the training bill unpriced

The memorandum names face blocking, character filters, C2PA credentials, watermarking and monitoring. It names no training-data licence, and that asymmetry sets the shipping checklist.

Publishers:dev.to

Reality

Evidence32
Adoption26
Hype gap+14
Incentives66
Confidence38
productOne report1 publisher

Rogue Studio ships an uncensored video model and prices permission from $29 to $299

A new platform sells uncensored AI video next to NanoBanana, Seedance and Kling 2.0, arguing restrictions now separate products. Its founders will not print their names.

Publishers:wired.com

Reality

Evidence34
Adoption12
Hype gap+42
Incentives76
Confidence44