{"schema":"postcutoff/event@1","as_of":"2026-11-13T13:49:00+01:00","url":"https://postcutoff.com/e/2026-09-15-stepfun-stepaudio-3/","md":"https://postcutoff.com/e/2026-09-15-stepfun-stepaudio-3/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"human editor","policy":"https://postcutoff.com/about/"},"license":null,"id":"2026-09-15-stepfun-stepaudio-3","date":"2026-09-15","date_precision":"day","short_title":"StepFun releases StepAudio 3 family","deck":"Its Realtime model tops Artificial Analysis full-duplex rankings","takeaway":"Chinese lab StepFun launched StepAudio 3, five audio models (Realtime, ASR Max, TTS, Gen, Music).","category":"model-release","category_label":"Model releases","importance":3,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":[{"n":1,"title":"StepFun on X - Introducing StepAudio 3","url":"https://x.com/StepFun_ai/status/2099916376274313630","type":"official","group":"primary","domain":"x.com"},{"n":2,"title":"StepFun audio models docs","url":"https://platform.stepfun.ai/docs/en/guides/models/audio","type":"docs","group":"primary","domain":"platform.stepfun.ai"},{"n":3,"title":"StepFun pricing","url":"https://platform.stepfun.ai/docs/en/pricing/details","type":"docs","group":"primary","domain":"platform.stepfun.ai"},{"n":4,"title":"StepAudio 3 Realtime Technical Report","url":"https://arxiv.org/abs/2609.14005","type":"paper","group":"primary","domain":"arxiv.org"},{"n":5,"title":"StepAudio 2.5 Realtime project page","url":"https://stepaudiollm.github.io/step-audio-2.5-realtime/","type":"official","group":"primary","domain":"stepaudiollm.github.io"},{"n":6,"title":"Decrypt - StepFun's voice AI topped every benchmark (StepAudio 2.5)","url":"https://decrypt.co/369013/stepfun-stepaudio-voice-ai-tops-benchmarks","type":"press","group":"press","domain":"decrypt.co"},{"n":7,"title":"Artificial Analysis on X - StepAudio 3 ASR #1 on AA-WER","url":"https://x.com/ArtificialAnlys/status/2102485740248842710","type":"discussion","group":"reactions","domain":"x.com"}],"official":5,"filed":"2026-09-29","updated":"2026-09-29","orgs":["StepFun"],"title":"StepFun releases StepAudio 3 family; its Realtime model tops Artificial Analysis full-duplex rankings","summary":"Chinese lab StepFun launched StepAudio 3, five audio models (Realtime, ASR Max, TTS, Gen, Music). StepAudio 3 Realtime, a \"think-while-speaking\" full-duplex voice model, ranked #1 on Artificial Analysis for Conversational Dynamics (98.9%) and Speech Reasoning (99.7%), and StepAudio 3 ASR ranked #1 on AA-WER (1.7%).","key_facts":["API ids: stepaudio-3-realtime-preview, stepaudio-3-chat-preview, stepaudio-3-asr-max, stepaudio-3-tts, stepaudio-3-gen-preview, stepaudio-3-music-preview","Realtime/Gen/Music free during preview; ASR Max $0.40/hour; TTS $0.36 per 10k characters","Realtime runs private reasoning in parallel with speech (Think-While-Speaking); 98.9 on Artificial Analysis Full-Duplex Bench","StepAudio 3 ASR 1.7% WER on AA-WER (StepAudio 2.5 ASR: 4.7%) per Artificial Analysis","Follows StepAudio 2.5 Realtime (2026-05-26): persona/role-play realtime model (zh/en) with million-scale persona augmentation and role-play RLHF; project page reports 80.41 human eval, 86.36 general dialogue, 79.80 spoken QA, 82.18 paralinguistics, first on all five of StepFun's own dimensions"],"key_numbers":[],"tags":["voice","speech","full-duplex","asr","tts","music","china"],"science":null,"body_md":"## What happened\n\nStepFun released a full audio stack at once and made the Realtime, Gen and Music models free during a preview period.\nThe Realtime model's technical report describes a listen-converse-think-act loop with \"Deep Perception\", \"Seamless Duplex\"\nand \"Think-While-Speaking\" components.\n\n## Why it matters\n\nA Chinese startup's voice model led a major independent leaderboard on conversational dynamics ahead of Western\nfrontier-lab voice models (GPT-Live-1 per StepFun's comparison), showing how fast full-duplex voice is commoditizing.\n\nLeaderboard positions are as of launch and come from StepFun's and Artificial Analysis's X posts.","disputed":[],"related":[{"id":"2026-09-23-qwen-audio-3-1","url":"https://postcutoff.com/e/2026-09-23-qwen-audio-3-1/","date":"2026-09-23","date_precision":"day","short_title":"Alibaba launches Qwen-Audio-3.1 five-model voice stack and cuts audio API prices up to 95%","deck":null,"takeaway":"Chinese labs (Alibaba, StepFun, ByteDance) now field voice-agent models that top or approach GPT-Live / Gemini Live on public leaderboards at a fraction of the price, turning real-time voice into a price war.","category":"model-release","category_label":"Model releases","importance":3,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":10,"official":8,"filed":"2026-09-29","updated":"2026-09-30","orgs":["Alibaba","Qwen"]}],"people":[],"posts":[],"videos":[],"models":[{"id":"stepaudio-3-asr-tts","name":"StepAudio 3 ASR Max / StepAudio 3 TTS","url":"https://postcutoff.com/m/stepaudio-3-asr-tts/"},{"id":"stepaudio-3-realtime","name":"StepAudio 3 Realtime","url":"https://postcutoff.com/m/stepaudio-3-realtime/"}],"changes":[{"date":"2026-09-29","type":"edit","text":"Added StepAudio 2.5 Realtime project page and its self-reported scores"},{"date":"2026-09-29","type":"filed","text":"Created"}],"provenance":{"agents":[{"model":"Claude Opus 5.5","maker":"Anthropic","tool":"Claude Code"}],"filed":"2026-09-29","run":null,"sources_read":null,"updated":"2026-09-29","human_review":null,"version":{"date":"2026-10-04"}},"gaps":[{"model_id":"gpt-6-astra","name":"GPT-6 Astra","cutoff":"2026-04","days_after":138,"in_training_data":false},{"model_id":"claude-opus-5-5","name":"Claude Opus 5.5","cutoff":"2026-06","days_after":77,"in_training_data":false},{"model_id":"gemini-4-hydrogen","name":"Gemini 4 Hydrogen","cutoff":"2026-03","days_after":168,"in_training_data":false},{"model_id":"grok-4-7","name":"Grok 4.7","cutoff":"2026-05","days_after":107,"in_training_data":false}],"short_url":null}