docs(livekit): update TTS for Lightning v3.1 Pro

[harshitajain165]

Summary

Updates the LiveKit integration docs to reflect the changes in livekit/agents#5799:

• Default model changed from lightning-v3.1 → lightning_v3.1_pro (underscore notation, new default in the plugin)
• Removed the <Note> block that said Lightning v3.1 Pro was unavailable in the plugin — it now ships in livekit-plugins-smallestai
• voice_id default documented as auto-resolved: meher for lightning_v3.1_pro, sophia for lightning_v3.1
• model parameter table updated with both model options and their capabilities (44.1 kHz Pro vs 217-voice standard)
• Language support clarified: Pro supports en, hi, auto only; standard supports 12 codes + auto
• sample_rate supported values corrected to 8000, 16000, 24000, 44100
• Notes bullet updated to name lightning_v3.1_pro as the new default

Test plan

• Verify rendered docs at https://docs.smallest.ai/waves/v-4-0-0/integrations/live-kit after merge

🤖 Generated with Claude Code

[github-actions]

🌿 Preview your docs: https://smallest-ai-preview-0d65c8c6-9410-48b2-9c30-96d7b4f0a2ac.docs.buildwithfern.com

[entelligence-ai-pr-reviews]

EntelligenceAI PR Summary

Updates LiveKit integration documentation with revised TTS defaults and expanded parameter details.

• Changed default TTS model from lightning-v3.1 to lightning_v3.1_pro
• Changed default voice from sophia to meher
• Removed <Note> block that previously flagged Pro model as unsupported by the plugin
• Added model-specific defaults, supported language codes, speed range (0.5–2.0), and valid sample rates (8000, 16000, 24000, 44100) to the parameter table

---

Confidence Score: 5/5 - Safe to Merge

Safe to merge — this documentation-only PR cleanly updates the LiveKit integration docs to reflect the new default TTS model (lightning_v3.1_pro), default voice (meher), and expanded parameter details including speed range and valid sample rates. The removal of the <Note> block flagging Pro model as unsupported is a meaningful accuracy improvement that aligns docs with current plugin capabilities. No logic, runtime, or security concerns exist in a pure documentation change of this nature, and the review found zero issues across all changed files.

Key Findings:

• The default model change from lightning-v3.1 to lightning_v3.1_pro and default voice change from sophia to meher are factual documentation updates with no code impact, and no reviewer raised accuracy concerns about these values.
• Addition of model-specific defaults, supported language codes, speed range (0.5–2.0), and valid sample rates (8000, 16000, 24000, 44100) expands the parameter table with concrete, actionable detail that improves developer experience.
• Removal of the <Note> block that previously flagged Pro model as unsupported correctly reflects current plugin support status, eliminating a potentially misleading warning for users.
• All heuristic checks passed with zero critical, significant, or medium issues identified, and 100% of changed files were reviewed with no unresolved comments.