A world made with
generous creators.

Peekawild uses original animal art and locally stored narration and nature sounds.

Animal models

All 104 animal models are original voxel artwork made for Peekawild.

Animal images

All 104 learning card images were generated for Peekawild. No Wikimedia Commons photographs are served in the game.

Movement and habitat sounds

Twelve shared cues were generated with ElevenLabs Sound Effects v2: hoofsteps, pawsteps, tiny feet, heavy steps, hops, small and large wing flaps, insect wing movement, leaf sliding, small lake waves, sea surf and river flow. These are AI-generated movement and habitat sounds, not animal vocal calls or field recordings.

Each discovery plays its cue, a brief pause, then Clara’s introduction. No celebration music, clapping or cheering plays. Clips were level matched, gently faded and encoded locally. Generation provenance and species mappings.

Spoken introductions

English introductions were generated with ElevenLabs eleven_v4, using Clara V — Warm & Articulate. All 61 English sentences were generated in one batch, then split and normalized locally. The Cantonese introductions came from a user-supplied Clara V batch and were also split locally. Mandarin, French, Japanese, and Korean animal introductions and habitat names were generated in timestamped ElevenLabs Clara V batches and split locally. The 30 added species have matching Clara V introductions generated in one additional timestamped batch per language. This is synthetic narration.

All game assets are served with the game. Playing does not send recordings or speech requests to an AI service.

Peekawild is made by LabLamb Works Limited.