Inworld Realtime TTS-2 tops voice leaderboard at 2.5x Eleven v3's speed, 1/5 price
ArtificialAnlys · x · 2026-09-03
Artificial Analysis' Controlled Voice leaderboard crowns Inworld Realtime TTS-2 at Elo 1123, ahead of Cartesia Sonic 3.6 (1119) and ElevenLabs Eleven v3 (1062). It generates 106 characters per second vs 40 for Eleven v3, and costs $20.8/M characters — a fifth of Eleven v3's $100 and less than half of Sonic 3.6's $49. Open-weights models including Breeze TTS 2, Mistral Voxtral TTS and Fish Audio S2 Pro land in the top 16. Voting is open in the Speech Arena, with clips in the Speech Explorer.
Related event: Inworld's Realtime TTS-2 Launches and Tops Voice Leaderboards(7 posts)→
More from Models
- Running Ling 3.1 Flash (560B MoE, 25B Active) Across 4 DGX Sparks via llama.cpp RPC — BerryBlush_Rogers691 · 2026-10-18
- Users press OpenAI on long-delayed Adult Mode a year after it was promised — ___Patrice___ · 2026-10-18
- Polymarket opens betting market on who releases the next Gemini Argon model — Polymarket · 2026-10-18
- Grok matter-of-factly lists which groups user was racist against, exempting Africans and Latin Americans — teortaxesTex · 2026-10-18
- A quick test for benchmaxxed demos: swap in Mirror's Edge and see if it generalizes — teortaxesTex · 2026-10-18
- Did OpenAI distill DeepSeek's gains? Benchmark shows OpenAI's Luna scoring just 38 — teortaxesTex · 2026-10-18