VibeVoice-1.5B is a Microsoft-originated open-source text-to-speech model for expressive, long-form, multi-speaker audio, reportedly capable of generating up to 90 minutes of speech. Its architecture uses continuous acoustic and semantic tokenizers running at 7.5 Hz to make long-sequence generation more computationally efficient. The supplied snippets do not independently substantiate the iPhone demonstration, its claimed ~2.2GB memory use, or the reported 1.28× real-time speed; the search summary instead says it is not near-real-time, so those performance claims remain conflicting and need reproduction.
2026-08-14T12:29:29Z
After repeated staleness checks, the promised iOS framework, runnable code, and independent benchmarks still have not appeared, so this demo-driven episode has faded without validation. Expire it and open a fresh episode if a testable artifact or external reproduction emerges.
2026-08-12T11:36:55Z
The staleness check confirms the episode remains dormant rather than advancing: no runnable iOS artifact or independent measurement has emerged to test the original demo. Revisit only if the promised xcframework, code, or an external reproduction appears.
2026-08-10T11:32:49Z
The staleness check and trivial score increase add no technical evidence; the iPhone performance claim remains a single first-party demo without a runnable framework or independent measurements. Keep the case dormant pending the promised xcframework, code, or external reproduction.
2026-08-08T11:23:19Z
Refreshed comments add only enthusiasm and speculative use cases, not a runnable iOS artifact or independent measurement. The iPhone performance claim remains uncorroborated; revisit only when the promised xcframework, code, or external benchmark appears.
2026-08-06T10:27:51Z
The attachment is null and adds no independent reproduction, runnable iOS artifact, or reconciled benchmark. This remains repetitive engagement-driven amplification; revisit only when the promised xcframework, code, or external measurements appear.
2026-08-06T09:26:27Z
The latest trigger is only a slight engagement decline, with no runnable artifact, independent reproduction, or reconciled performance data. Keep the case dormant and stop engagement-driven checks until the promised xcframework, code, or an external benchmark appears.
2026-08-06T00:27:57Z
The new attachment is null and adds no runnable artifact, independent reproduction, or reconciled benchmark; this is continued engagement-driven amplification rather than evidence. Keep the case dormant until code, the promised xcframework, or external measurements appear.
2026-08-05T21:27:32Z
The new attachment adds no runnable iOS artifact, independent reproduction, or reconciled benchmark; it is continued amplification of the same first-party demo. Leave the case dormant until the promised xcframework, code, or external measurements appear.
2026-08-05T18:29:57Z
The trigger adds no identifiable artifact or independent evidence, so the case remains repetitive amplification of the original first-party demo. Defer further review until the promised xcframework, runnable code, or an external benchmark appears.
2026-08-05T17:29:25Z
The latest trigger adds no artifact or independent evidence, only another re-observation of the original demo. Keep the claim dormant until the promised xcframework, runnable code, or external benchmarks appear.
2026-08-05T15:25:13Z
The attachment is another re-observation of the original first-party demo, not an independent reproduction or runnable iOS release. The case remains dormant until the promised xcframework, code, or external benchmarks can resolve the conflicting performance claims.
2026-08-05T13:29:38Z
The trigger contains no substantive new evidence beyond the same first-party demo, so the deployment claim remains uncorroborated and its conflicting speed figures unresolved. Stop engagement-driven checks and wait for the promised xcframework, runnable code, or an independent benchmark.
2026-08-05T11:28:33Z
No substantive new evidence has appeared; the case remains repetitive amplification of the same first-party demo, without a runnable iOS artifact or independent benchmark. Keep it dormant until the promised xcframework, code, or external reproduction arrives.
2026-08-05T09:26:07Z
The attached item adds no runnable artifact, independent reproduction, or reconciled performance data; it is repetitive amplification of the original demo. Keep the case dormant until the promised xcframework, code, or external benchmarks appear.
2026-08-05T08:28:17Z
The new attachment adds no independent evidence and remains another observation of the same first-party demo. Pause engagement-driven review until a runnable xcframework, code release, or external benchmark can test the conflicting performance claims.
2026-08-05T07:22:33Z
No substantive new evidence is present: the case still rests on the same first-party demo without a runnable iOS artifact, independent reproduction, or reconciled performance figures. Repeated engagement-driven observations no longer merit frequent review; wait for the xcframework, code, or external benchmarks.
2026-08-05T06:26:57Z
The latest attachment adds no independent reproduction, released xcframework, or reconciled latency and memory measurements; this remains repetitive amplification of one first-party demo. Further repricing should wait for runnable iOS artifacts or external benchmarks.
2026-08-05T05:23:59Z
The new attachment still provides no independent reproduction, released xcframework, or reconciled performance measurement; it is repetitive amplification of the same first-party demo. The iPhone deployment claim remains plausible but uncorroborated and internally inconsistent.
2026-08-05T04:23:46Z
The attached evidence adds no independent reproduction or released iOS framework; it still resolves to the same first-party demonstration with conflicting speed figures. The claim remains a bounded but uncorroborated deployment result awaiting code and external benchmarks.
2026-08-05T03:24:59Z
The added observation still traces to the same first-party demo, while engagement barely moved and no code, xcframework, or independent benchmark has appeared. The bounded iPhone performance claim remains plausible but internally inconsistent and uncorroborated.
2026-08-05T02:27:08Z
grounded: known/medium — The radar already tracks this validation question in “Independent benchmarks will determine whether audio.cpp 0.4 delivers practical real-time local TTS and ASR
2026-08-05T02:24:53Z
origin walked (codex/luna, conf 0.77): anchor reddit.post.1vfuawi -> echo.other.efc046d0a4 by None
2026-08-05T02:21:31Z
case created — The concrete on-device audio demo makes a bounded deployment claim, but code and independent measurements are not yet available.