MAKEUGC PLATFORM API · DOCS V2
Limits & credits
Understand API credit units, model-specific input limits and practical guidance for production request volume.
Credit units
estimated_credits,
credits_charged, credits_used and template
credit use API units. A value of 0.1
corresponds to 10 credits on the website. Preserve the raw response
value in your integration.
Cost depends on the feature, model and inputs. For quote-based models, returned charge fields take precedence over an estimate. Do not send prices, free-generation flags or credit overrides in a public API request.
Generation costs
| Generation | API credit units | Displayed credits |
|---|---|---|
| Talking actor / custom avatar / Nova V2 | 1 | 100 |
| Omi-Human | 1 under 15s; 2 for 15–30s | 100 / 200 |
| Product-in-Hand Veo | 1 Fast or 2 Pro per scene | 100 / 200 per scene |
| GPT Image 2 / Nano Banana Pro / FLUX.2 Pro / Grok Imagine Image (legacy enterprise IDs) | 0.1 per generation | 10 |
| Enterprise Veo 3.1 | 0.5 Lite; 1 Fast; 2 Standard | 50 / 100 / 200 |
| Content Library / Shopify / Shoppable | Template or job dependent | API units × 100 |
| Grok Imagine 1.5 / Gemini Omni | Quoted from inputs | Returned charge × 100 |
| Additional image models / MiniMax / WAN 3.0 | Depends on model, output settings and source media | Returned charge × 100 when provided |
| WAN Animate | 1 per started 30s of source video | 100 per started 30s |
Seedance per-second rates
| Model | 480p | 720p | 1080p | 4k |
|---|---|---|---|---|
| 2.0 Pro | 0.05 | 0.11 | 0.25 | 0.52 |
| 2.0 Fast | 0.04 | 0.09 | — | — |
| 2.5 | 0.07 | 0.16 | 0.38 | — |
Rates are API credit units per output second. Auto duration is estimated
at 15 seconds for 2.0 or 30 seconds for 2.5. The dedicated endpoint
returns estimated_credits and, when available,
credits_charged. A ten-second Seedance 2.5 request at 720p
estimates 1.6 API units (160 displayed credits).
Refund and adjustment fields are model-specific. Grok/Gemini responses
may include refund_result.credits_refunded or
credit_adjustment.credits_refunded. Read returned values;
do not assume a universal refund policy.
Input limits
- Talking-actor script: 1,500 characters by default. Existing audio: at most 120 seconds.
- Custom avatar uploads: at most 10 MiB each for image and voice.
- Product-in-Hand scripts: Nova V2/Omi-Human up to 2,000 characters. Audio: 90 seconds Nova V2; 30 seconds Omi-Human. Veo: 1–4 scenes.
- Seedance prompt: 15,000 Unicode characters. Dedicated 2.0: 4–15 seconds; 2.5: 4–30 seconds; both accept -1 for Auto.
- Grok Imagine 1.5 prompt: 4,096 characters; duration 1–15 seconds. Reference mode: 1–7 images.
- Gemini Omni prompt: 20,000 characters; generation duration 3–10 seconds. Edit uses source duration. Reference mode: up to 10 images and 3 videos, each video at most 3 seconds.
- Other enterprise image/Veo prompts: up to 2,000 characters. Image edits: up to 2 source URLs. Veo duration: 8 seconds.
- Template inputs and supported media vary. Read the individual model page or template’s require_fields.
Request volume
A universal requests-per-minute quota is not published by these routes. Capacity and generation limits can vary by account, model and service availability. Do not assume that an unlimited web plan provides unlimited API generation.
Start with low concurrency, cache discovery lists appropriately, and poll each active job every 5–10 seconds. On 429, reduce concurrency and honor Retry-After if provided. Contact support to confirm capacity for a high-volume integration.