Pika

Pika (/media/pika) is a playful AI video platform best known for Pikaffects, one-tap surreal effects that transform a photo into shareable video. It also supports standard text-to-video and image-to-video generation.

Where to configure

Pika's official API is hosted on fal.ai, so the studio rides the fal adapter. Configure your fal.ai API key in Settings → Media. See Settings for provider setup.

Tabs / operations

Tab	Operation	Key fields	Output
Text → Video	Video	Prompt, negative prompt, aspect ratio, resolution (720p/1080p), duration (5/10s), optional seed	MP4 video clip
Image → Video	Video	Source image, prompt, negative prompt, resolution, duration, optional seed	MP4 video clip
Pikaffects	Video	Source image, effect (e.g. Cake-ify, Melt, Inflate), optional prompt, optional seed	MP4 video with the chosen one-click VFX

Generation flow

Video jobs are submitted to the fal.ai queue and tracked in the Render Queue. When complete, the artifact is saved to /files and can be Edited in Designer or Posted. See Media Studios overview for the shared flow.

Caveats

• The studio uses the fal.ai key and queue infrastructure; Pika itself is identified by the full fal endpoint id (e.g. fal-ai/pika/v2.2/text-to-video).
• fal job ids are namespaced as <model>::<request_id> so the poll path can route back to the correct model endpoint.
• Video completions rely on polling; there is no provider webhook.
• The model field is lifted out of input by the shared Studio Kit backend; all other native fal/Pika parameters ride through unchanged.
• Pikaffects do not require a prompt; the effect alone drives the transformation, with an optional prompt for extra guidance.

---

Verified against v1.0.0