Starter Pack100 free points on sign upChat, image, speech. Expires in 7 days. One per verified email and device.
Category packs: Video, Image, Audio, Text
Pick a pack for what you make most. Each size costs the same as the all-purpose pack and the points work anywhere, so nothing gets stranded. Counts are estimates from a cheapest, a middle and a premium model.
| Pack | Price | What it buys (estimates) |
|---|---|---|
| Video Pack | $20 | 56 budget 10s clips; 20 mid 10s clips; 4 premium 10s clips with audio |
| Video Pack | $50 | 144 budget 10s clips; 52 mid 10s clips; 11 premium 10s clips with audio |
| Video Pack | $100 | 300 budget 10s clips; 108 mid 10s clips; 24 premium 10s clips with audio |
| Image Pack | $20 | 5,100 budget images; 255 standard images (1K); 42 premium 4K images |
| Image Pack | $50 | 13,000 budget images; 650 standard images (1K); 108 premium 4K images |
| Image Pack | $100 | 27,000 budget images; 1,350 standard images (1K); 225 premium 4K images |
| Audio Pack | $20 | 1,133 10s premium voice clips; 485 10s dialogue voice clips; 85 music tracks; 5,100 minutes of transcription |
| Audio Pack | $50 | 2,888 10s premium voice clips; 1,238 10s dialogue voice clips; 216 music tracks; 13,000 minutes of transcription |
| Audio Pack | $100 | 6,000 10s premium voice clips; 2,571 10s dialogue voice clips; 450 music tracks; 27,000 minutes of transcription |
| Text Pack | $20 | 5,100 small-model messages; 1,020 mid-model messages; 364 top-model messages |
| Text Pack | $50 | 13,000 small-model messages; 2,600 mid-model messages; 928 top-model messages |
| Text Pack | $100 | 27,000 small-model messages; 5,400 mid-model messages; 1,928 top-model messages |
Checkout opens with the rest of the store.
Order: inside Video, Image and Intelligence, models are sorted by real public usage as of Oct 19, 2026. Video and image: share of generations on Magnific and Freepik, Nov 2025 to Aug 2026 (magnific.com/blog/ai-video-models, ai-image-models). Intelligence: tokens processed on OpenRouter, week to Oct 19, 2026 (openrouter.ai/rankings). Models with no published data follow the ranked ones in their old order. Voice, music, speech to text and data tools are not ranked: no public usage data found.
Stars (text models only): 1 to 5 by quintile of the public Artificial Analysis intelligence index, read from a public model list on Oct 19, 2026. No stars means no public data, not a low score. Favorites: no public count exists, so this is our own count, one favorite per key, blank until real. Customer ratings will replace the stars once they exist.
How to read the numbers: video lines count 10-second videos; models that sell fixed 5s clips are counted as clips joined to reach 10s. Voice lines count 10-second clips of speech (assumed about 150 characters). Speech to text counts minutes of audio. Image lines count one image at 1K (or the cheapest size if no 1K). The variant used is named under each model. each cell is the estimated number of uses that tier buys (tier points divided by points per use). Counts are shown for every tier. Policy note: new buyers may have video limited during their first 7 days. A 0 means one use costs more points than that tier holds.