Avatar vision & voice options (free vs bring-your-own-key)
The avatar needs two AI services: vision (understanding what's on screen) and voice (speaking its lines). Both have a free built-in tier and a bring-your-own-key upgrade, configured in the Avatar Studio:
- Vision โ Preset (free): built in, no key needed. BYOK: OpenAI (GPT-4o) or Google Gemini keys give noticeably better game understanding; each provider has its own pricing and free tiers.
- Voice โ Browser TTS (free): works out of the box. BYOK: an ElevenLabs key unlocks natural, characterful voices.
Key safety
Keys are stored securely and never shown again after saving โ you'll only ever see a masked version like sk-โฆxY. To remove a key, enter __clear__ in the key field and save.
Your keys are billed by their provider, not by StreamTracker โ set spending limits in your OpenAI/Google/ElevenLabs account if you're worried about usage.