AI Models and Availability
Disclosed AI video models (ByteDance Seedance 2.0 Mini and Fast via Kie), credit costs, input limits, and output expectations for Talk to Me Doja AI.
Last updated: 2026-11-18
1. AI Model Developer and API Service
Talk to Me Doja AI (https://talktomedoja.net) generates 12-second vertical 9:16 dance videos using Seedance 2.0 video models developed by ByteDance, accessed through the Kie API service (https://api.kie.ai/api/v1).
- Model Developer: ByteDance (
bytedance/seedance-2-miniandbytedance/seedance-2-fast). - API Intermediary: Kie (
https://api.kie.ai/api/v1). - Independence Notice: Naming ByteDance and Kie identifies the underlying technology powering video inference and does not imply any official partnership, sponsorship, or endorsement. Talk to Me Doja AI is also independent of Doja Cat, Gucci Mane, Kemosabe Records, RCA Records, and Yamada-kun and the Seven Witches.
2. Available Tiers, Models, and Credit Pricing
Every tier uses the same five-beat choreography prompt for a 12-second vertical 9:16 MP4. Timing and motion vary with each generation:
- Fast · 480p Standard (
15 credits): Usesbytedance/seedance-2-miniat480presolution (estimated render time ~3–5 minutes; approx.$4.99in Starter or$3.75in Studio). - Fast · 720p HD (
22 credits): Usesbytedance/seedance-2-miniat720presolution (estimated render time ~4–6 minutes; approx.$6.35in Creator or$5.50in Studio). - Premium · 480p Standard (
30 credits): Usesbytedance/seedance-2-fastat480presolution (estimated render time ~4–6 minutes; approx.$8.66in Creator or$7.50in Studio). - Premium · 720p HD (
45 credits): Usesbytedance/seedance-2-fastat720presolution (estimated render time ~5–8 minutes; approx.$12.99in Creator or$11.25in Studio).
Note: 1080p is not offered. App credits come in one-time packs (Starter $4.99 / 15 credits, Creator $12.99 / 45 credits, Studio $29.99 / 120 credits). There are no subscriptions, and purchased credits never expire. See our Pricing page for full pack details.
3. Input Specifications and Audio Track
- Photo Input: One clear, front-facing adult portrait (
JPG,PNG, orWebP, under 20 MB, minimum 300 × 300 px). Half-body or waist-up framing works best because the model preserves your outfit. - Audio option: You can ask the model to create a soundtrack or switch audio off before generating. The app does not include commercial recordings by Doja Cat or Gucci Mane.
4. Safety Screening, Processing, and Limitations
- Consent & Safety Flags: Generation requires a signed-in account and explicit confirmation of the mandatory photo-rights checkbox. Every job sends
nsfw_checker: trueto Kie along with the fixed template prompt and your uploaded photo. See our Acceptable Use Policy for prohibited content and current moderation limits. - Automatic credit refunds: If a charged task fails or is rejected in pre-delivery review, its full credit charge is returned. Submitted tasks cannot be canceled once accepted by the provider.
- Output expectations: Likeness, hand rendering and outfit continuity vary by photo. Our Examples page shows a Fast 480p render made from the displayed sample portrait.
- Retention: Source portraits on our storage are scheduled to expire after 30 days. See our Privacy Policy for media links and third-party processing.