Phonic-1 tops DIAL speech-to-speech benchmark on naturalness and content, 4.04/5
graceisford · x · 2026-11-10
DIAL added two speech-to-speech models to its benchmark: Phonic-1 from Phonic and Realtime v1 mini from ElevenLabs, evaluated via live multi-turn conversations with human raters.
- Overall preference: Phonic-1 ranks #2 at 4.04/5; Realtime v1 mini ranks #12 at 3.60/5
- The poster says Phonic-1 is #1 on naturalness and content, outperforming leading frontier-lab models
- Rankings can be broken down across Consumer vs Business conversations, with separate scores for naturalness, audio quality, content and user-flagged issues
More from Multimodal
- Local pipeline showcase: MiniMax H3 refMod generations from reference image to 4K upscale — edisson75 · 2026-11-13
- Gaussian splats may be the best way to capture how 3D displays like LKG Portrait look IRL — nptacek · 2026-11-13
- AI-generated 'not canon' clip of Schwarzenegger driving with Michael Jordan — alanskimp · 2026-11-13
- H3-Image API gotcha: server keeps billing on client timeout — set max_retries=0 — xiaohu · 2026-11-13
- MiniMax open-sources H3-Image: native 16MP output, up to 16 reference images — xiaohu · 2026-11-13
- Daily AI painting made with GPT-6 Images 2.6 shows painterly winter village — DeryaTR_ · 2026-11-13