Independent testing will determine whether ClearVoice can run OmniVoice multilingual text-to-speech and voice cloning fully offline on supported iPhones and iPads with acceptable quality, latency, and peak memory use.

state: expiredheat: lowuncertainty: highconvergesscott: mediumlocal-inference text-to-speech on-device-aiClearVoiceOmniVoice

What is this?

ClearVoice is presented as an iOS/macOS app that claims to run the OmniVoice multilingual text-to-speech and zero-shot voice-cloning model fully offline; an earlier announcement attributes the app claim to Rory Clear and OmniVoice to XiaomiMiMo. The supplied OmniVoice repository and model-page snippets describe support for 600+ languages, voice cloning, voice design, and fast inference, while one third-party benchmark reports roughly 2 GiB of peak VRAM on unspecified hardware. None of the supplied search snippets independently tests ClearVoice on an iPhone or iPad, so device compatibility, offline operation, output quality, latency, and peak memory on supported Apple mobile hardware remain unverified here.

Why it matters to Scott

ClearVoice independently pursues the same offline speech and voice-cloning territory as Scott’s local speech-engine laboratory, while its unresolved iPhone quality, latency, and memory claims map directly to his hardware-aware inference and capability-audit approach. Validation could identify a practical native-mobile deployment path or benchmark for his voice stack, but the supplied evidence is still only a product claim.

dev:project.audiodev:concept.hardware-aware-local-inferenceip:concept.capability-auditdev:technology.f5-ttsradar:concept.local-ttsradar:vibevoice-iphone-local-inferenceradar:llama-cpp-qwen3-tts-voice-cloningradar:concept.model-evaluation

queries asked of Scott's wikis

on-device inference economics and privacy
local model benchmarking on Apple silicon
offline voice interfaces and speech tooling
mobile inference memory and latency constraints
open-weight models in native iOS apps
voice cloning safety and consent patterns

Measured heat

no measured readings yet — the hourly heat pass fills this in

How the heat travelled

no chain yet — the hourly chain pass fills this in

Evidence (3) — ⭐ canonical anchor

source	object	author	score	comments
🟧 hn	Show HN: ClearVoice TTS, SOTA voice cloning model running offline, in an iOS app	roryclear	1	0
🟧 echo.x ⭐	Rory Clear’s earlier primary announcement says: “You can now run a decent voice cloning model (OmniVoice @XiaomiMiMo) in an iOS or MacOS app	Rory Clear	—	—
🟠 reddit	iPhone Local TTS EPUB Reading - Audiobookify LocalLLaMA	therysin	48	52

Interpretation history

2026-08-26T21:25:46Z
Repeated refreshes have produced only adjacent mobile-TTS discussion, with no independent ClearVoice test or indication that one is imminent; the validation episode has faded and can be reopened if device benchmarks appear.
2026-08-24T20:47:20Z
Another Audiobookify comment refresh adds only adjacent product interest, not an independent ClearVoice or OmniVoice test. The case remains a developer-reported artifact with mobile quality, latency, compatibility, and peak memory unresolved.
2026-08-24T18:27:54Z
The refreshed Audiobookify discussion remains adjacent ecosystem chatter, not an independent ClearVoice or OmniVoice test. Repeated comment updates add no case-specific validation, so the hypothesis and its high uncertainty are unchanged.
2026-08-24T17:27:47Z
The refreshed discussion is repetitive ecosystem interest rather than new case-specific evidence. ClearVoice still lacks an independent iPhone or iPad test of offline operation, cloning quality, latency, compatibility, and peak memory.
2026-08-24T16:31:08Z
The refreshed comments remain general product feedback and adjacent mobile-TTS interest, adding no independent ClearVoice test or OmniVoice device benchmark. The case still rests on the developer’s artifact and claims, with quality, latency, compatibility, and peak memory unresolved.
2026-08-24T15:27:07Z
The refreshed discussion adds another adjacent Kokoro/WebGPU implementation but no independent ClearVoice test or OmniVoice benchmark; it reinforces general mobile TTS feasibility without changing the case-specific uncertainty.
2026-08-24T14:29:44Z
Audiobookify modestly strengthens the feasibility prior for offline iPhone TTS, but it uses different models and does not validate ClearVoice, OmniVoice voice cloning, or the claimed device performance. The case remains a developer-reported artifact awaiting independent benchmarks.
2026-08-24T14:22:40Z
evidence attached: reddit.post.1vx2yg4 — A working iPhone app using hand-optimized local TTS models provides practical ecosystem evidence for fully offline on-device speech generation.
2026-08-22T16:31:13Z
No new testing or implementation evidence has appeared; the case remains a developer-reported mobile artifact awaiting independent validation of offline operation, quality, latency, compatibility, and peak memory.
2026-08-22T16:28:15Z
grounded: converges/medium — ClearVoice independently pursues the same offline speech and voice-cloning territory as Scott’s local speech-engine laboratory, while its unresolved iPhone qual
2026-08-22T16:25:57Z
origin walked (codex/luna, conf 0.97): anchor hn.story.49400671 -> echo.x.110d38aa7a by Rory Clear
2026-08-22T16:24:50Z
case created — The released iOS app is a usable on-device inference artifact, but its quality, performance, and device compatibility remain independently unvalidated.

Decision trace

08-27 07:25expireRepeated refreshes have produced only adjacent mobile-TTS discussion, with no independent ClearVoice test or indication that one is imminent; the validation episode has faded and can be reopened if de
08-27 07:25alert_silentThe new delta is only refreshed discussion about a different offline TTS app and adds no consequential evidence about ClearVoice, OmniVoice cloning quality, latency, compatibility, or memory use.
08-27 07:25alert_routeThe new delta is only refreshed discussion about a different offline TTS app and adds no consequential evidence about ClearVoice, OmniVoice cloning quality, latency, compatibility, or memory use.
08-25 06:47repriceAnother Audiobookify comment refresh adds only adjacent product interest, not an independent ClearVoice or OmniVoice test. The case remains a developer-reported artifact with mobile quality, latency,
08-25 06:47alert_silentNo consequential new fact occurred; refreshed comments about a different offline TTS app do not validate ClearVoice and can wait for routine review.
08-25 06:47alert_routeNo consequential new fact occurred; refreshed comments about a different offline TTS app do not validate ClearVoice and can wait for routine review.
08-25 06:21sensor_dirtycomment_update
08-25 04:27repriceThe refreshed Audiobookify discussion remains adjacent ecosystem chatter, not an independent ClearVoice or OmniVoice test. Repeated comment updates add no case-specific validation, so the hypothesis a
08-25 04:27alert_silentNo consequential new fact emerged; general interest in another offline iPhone TTS app does not validate ClearVoice’s offline operation, cloning quality, latency, compatibility, or memory use.
08-25 04:27alert_routeNo consequential new fact emerged; general interest in another offline iPhone TTS app does not validate ClearVoice’s offline operation, cloning quality, latency, compatibility, or memory use.
08-25 04:21sensor_dirtycomment_update
08-25 03:27repriceThe refreshed discussion is repetitive ecosystem interest rather than new case-specific evidence. ClearVoice still lacks an independent iPhone or iPad test of offline operation, cloning quality, laten
08-25 03:27alert_silentNo consequential new fact emerged; refreshed Audiobookify comments neither test ClearVoice nor validate OmniVoice performance, so this can wait for routine review.
08-25 03:27alert_routeNo consequential new fact emerged; refreshed Audiobookify comments neither test ClearVoice nor validate OmniVoice performance, so this can wait for routine review.
08-25 03:21sensor_dirtycomment_update
08-25 02:31repriceThe refreshed comments remain general product feedback and adjacent mobile-TTS interest, adding no independent ClearVoice test or OmniVoice device benchmark. The case still rests on the developer’s ar
08-25 02:31alert_silentNo consequential new fact has emerged; repetitive discussion around Audiobookify does not validate ClearVoice or alter Scott’s decisions, so routine review is sufficient.
08-25 02:31alert_routeNo consequential new fact has emerged; repetitive discussion around Audiobookify does not validate ClearVoice or alter Scott’s decisions, so routine review is sufficient.
08-25 02:21sensor_dirtycomment_update
08-25 01:27repriceThe refreshed discussion adds another adjacent Kokoro/WebGPU implementation but no independent ClearVoice test or OmniVoice benchmark; it reinforces general mobile TTS feasibility without changing the
08-25 01:27alert_silentThe new comment is ecosystem corroboration already implicit in the attached Audiobookify evidence, not validation of ClearVoice’s offline operation, cloning quality, latency, compatibility, or memory
08-25 01:27alert_routeThe new comment is ecosystem corroboration already implicit in the attached Audiobookify evidence, not validation of ClearVoice’s offline operation, cloning quality, latency, compatibility, or memory
08-25 01:21sensor_dirtycomment_update
08-25 00:29repriceAudiobookify modestly strengthens the feasibility prior for offline iPhone TTS, but it uses different models and does not validate ClearVoice, OmniVoice voice cloning, or the claimed device performanc
08-25 00:29alert_silentThe adjacent implementation is useful ecosystem evidence but does not materially resolve the ClearVoice-specific questions, so it can wait for the normal briefing or independent device testing.
08-25 00:29alert_routeThe adjacent implementation is useful ecosystem evidence but does not materially resolve the ClearVoice-specific questions, so it can wait for the normal briefing or independent device testing.
08-25 00:23alert_silentAudiobookify is a substantive adjacent implementation using hand-optimized Core ML conversions of Kokoro and Supertonic 3 for offline iPhone TTS, but it neither validates ClearVoice’s OmniVoice voice-
08-25 00:23surface_candidateAudiobookify is a substantive adjacent implementation using hand-optimized Core ML conversions of Kokoro and Supertonic 3 for offline iPhone TTS, but it neither validates ClearVoice’s OmniVoice voice-
08-25 00:23alert_routeAudiobookify is a substantive adjacent implementation using hand-optimized Core ML conversions of Kokoro and Supertonic 3 for offline iPhone TTS, but it neither validates ClearVoice’s OmniVoice voice-
08-25 00:22attachA working iPhone app using hand-optimized local TTS models provides practical ecosystem evidence for fully offline on-device speech generation.
08-25 00:22propose_attachA working iPhone app using hand-optimized local TTS models provides practical ecosystem evidence for fully offline on-device speech generation.
08-23 02:31repriceNo new testing or implementation evidence has appeared; the case remains a developer-reported mobile artifact awaiting independent validation of offline operation, quality, latency, compatibility, and
08-23 02:31alert_silentThis is an unchanged reobservation with no consequential new delta, so it can wait for independent device testing or the normal briefing.
08-23 02:31alert_routeThis is an unchanged reobservation with no consequential new delta, so it can wait for independent device testing or the normal briefing.
08-23 02:28alert_silentClearVoice appears to be an available native iOS/macOS implementation of offline OmniVoice voice cloning, but the only performance evidence is the developer’s own report, the announcement is not clear
08-23 02:28surface_candidateClearVoice appears to be an available native iOS/macOS implementation of offline OmniVoice voice cloning, but the only performance evidence is the developer’s own report, the announcement is not clear
08-23 02:28alert_routeClearVoice appears to be an available native iOS/macOS implementation of offline OmniVoice voice cloning, but the only performance evidence is the developer’s own report, the announcement is not clear
08-23 02:28groundClearVoice independently pursues the same offline speech and voice-cloning territory as Scott’s local speech-engine laboratory, while its unresolved iPhone quality, latency, and memory claims map dire
08-23 02:25promote_anchororigin walk conf 0.97
08-23 02:24createThe released iOS app is a usable on-device inference artifact, but its quality, performance, and device compatibility remain independently unvalidated.