Fish's 'Full Duplex' Voice Model Criticized: Same Pipeline Issues, Plus Paid Twitter Hype
realmrfakename · x · 2026-11-11
Developer realmrfakename criticized Fish's newly released 'full duplex' voice model in hands-on testing: despite the full-duplex claim, it shows all the issues of an ASR-LLM-TTS pipeline, with turn-taking only slightly better and no real sense of 'knowing when to shut up.'
He also called out the product's heavy reliance on paid Twitter promotion to hype it up (sharing an apparent UI bug as evidence and noting it's not a paid tweet), said Drama 3 wasn't great either, and is bearish on Fish overall.
Related event: Fish Audio's Full-Duplex Voice Model Slammed in Hands-On Tests(2 posts)→
More from Models
- Researcher lists three explanations for "model nerfing," one intentional — gleech · 2026-11-11
- GPT-4-turbo outputs shorter completions when told it's December, N=477 test — gleech · 2026-11-11
- Fable 5.2 Early Access Test: Video Editing Nearly One-Shot, Jump Cuts Sharply Reduced — FinanceYF5 · 2026-11-11
- Fable 5.2 Manga Test Falls Short: 12 Comics With Stiff Dialogue and Broken Anatomy — FinanceYF5 · 2026-11-11
- vLLM Upgrades EPD Disaggregation for Multimodal Serving: Lower TTFT, Higher Per-GPU Throughput — vllm_project · 2026-11-11
- Fable 5.2 takes over an existing Three.js/WebGPU hockey game to boost performance and add features — FinanceYF5 · 2026-11-11