Launching today

SantaPhoto
AI Christmas portraits, cards and avatars from your photos.
2 followers
AI Christmas portraits, cards and avatars from your photos.
2 followers
Create Christmas portraits, cards and avatars from your photos. Choose from 48 presets, combine reference photos, and generate with free sign-up credits or one-time packs. No subscription.



Free Options
Launch Team

TinesThe single, secure environment for agents, apps, and automations.
Promoted
Hi , we built SantaPhoto (https://santaphoto.app), a browser tool for creating Christmas portraits, holiday cards, and seasonal avatars from reference photos of individuals, couples, families, and pets.
Upload photos, pick from 48 presets across three modes (cozy fireplaces, matching pajamas, snowy cabins, claymation or illustrated cards, pet portraits), and render an image in 30 to 40 seconds.
Photo and card modes combine 1 to 4 separate reference photos into one scene; card mode adds a short greeting, and the avatar maker uses one photo.
Two ways to try it:
1. No account: browse and download avatars at https://santaphoto.app/christmas... (or card templates at https://santaphoto.app/christmas...), and preview all presets and before/after pairs on the home page.
2. Personalized generation: requires sign-in and is metered, not unlimited or free. Sign-up grants 6 free credits (1K costs 3, 2K costs 5, 4K costs 8), followed by optional one-time packs ($4.99 to $19.99, valid a year, no subscription). Failed renders refund automatically.
Even while building this, we have mixed feelings about synthetic family photos. Much of the time they feel hollow and meaningless: a generated living room is not a real memory, and nothing replaces sitting on the same couch. Still, when distance, work, or circumstances keep relatives apart over Christmas, merging separate phone shots into one frame is a reluctant fallback for families who cannot gather in person. On the practical side, most AI photo apps push weekly or monthly subscriptions, which makes little sense for something used once in December, so we stuck to one-time packs and self-serve deletion (deleting a creation immediately erases both the output and unreferenced source uploads from our storage).
We did not train custom image weights. Synthesis runs through OpenAI's GPT Image API for photorealistic presets and a secondary provider for claymation and illustrated presets, with automatic failover on timeouts or errors. What we built ourselves is the multi-photo prompt assembly, pre-generation moderation gate, async job and credit pipeline, a read-only MCP server (https://santaphoto.app/mcp), and a 9-locale web app built with TanStack Start (React 19, Vite, Nitro) on Cloudflare Workers, R2 object storage, and Paraglide JS. Source photos are resized client-side to ~2048px before upload.
The trickiest trade-off was keeping subject counts accurate without breaking multi-photo merges, so we ran 85 test renders and published every input and output (including failures) at https://santaphoto.app/blog/ai-c.... Merging 2 to 4 separate photos worked reliably and pet markings held up, but on solo uploads, group-worded presets ("Matching Pajamas", "Ugly Sweater Party") tried to fill the scene by inventing extra strangers or a dog in 4 of 10 single-person tests. Tightening subject-count rules in the server prompts fixed those solo retests while preserving real pets and group members, though dim or angled phone photos still vary more than our clean test portraits.
If you try it, feedback is welcome on:
- Multi-photo compositing with real phone photos (lighting or angle mismatches, person + pet combinations) and where likeness breaks down.
- Whether the boundary between the no-login gallery and metered generation is clear before uploading.
- Any mobile upload/resize bugs or rough translations across the 9 locales.