Vibe

What is Vibe?

Best suited for

• A creator, student, interviewer, independent professional or small-shop owner who wants local transcription from authorized recordings and can review names, numbers, negations and subtitle timing.
• A pilot focused on small-shop voice note transcription with exact quantities, budget and negation, using authorized local audio/video recordings, a selected compatible speech model and fixed transcription settings. The recorded native pilot used a 14.125-second local Microsoft David Desktop synthetic English shop note and an exact ten-second zero-sample WAV, both 16 kHz mono signed 16-bit PCM with frozen hashes, and one pinned ggml-tiny.bin model. Original case wording from the pre-resource planning state is preserved; a separate pre-inference contract supplies the actual bytes.

Not suited for

• A workflow that depends on the following request without the stated input, review or permissions: Run the same complete official Vibe 3.2.2 bundled-server 0.6.10 native transcribe route and exact pinned model/arguments as the primary, now using only the frozen distinct silence-input.wav from a separate fresh task-owned directory. Preserve the unedited actual empty transcript or native error. Do not add VAD, reuse a prior transcript, supply any prompt, retry with a new model or fill the silence with an explanation.
• Vibe is a specialist transcription tool. The advertised HTTP API or agent skills do not establish autonomous orders, publishing or business workflows.
• Speech recognition can change quantities, names, dates and negations or hallucinate text from silence. Human review is required; no broad transcription accuracy claim was reproduced.

Capabilities, with sources

• 01The official README advertises on-device audio/video transcription, batch input, microphone/system-audio features and Windows/macOS/Linux support.Official vendor statement · checked 2026-10-02Source ↗
• 02The official README lists SRT, VTT, TXT, HTML, PDF, JSON and DOCX output formats.Official vendor statement · checked 2026-10-02Source ↗
• 03Transcript summaries and analysis can use optional Claude API or Ollama; local transcription claims do not certify every network-connected feature.Official vendor statement · checked 2026-10-02Source ↗
• 04The fixed software license is MIT with Copyright (c) 2024 thewh1teagle.Official vendor statement · checked 2026-10-02Source ↗
• 05The desktop application bundles a native vibe-server process and communicates over local HTTP; on macOS/Windows it also bundles an FFmpeg helper.Official vendor statement · checked 2026-10-02Source ↗
• 06The fixed desktop CLI forwards arguments to the bundled server and includes lifecycle analytics calls; actual configuration and network behavior have not been inspected.Official vendor statement · checked 2026-10-02Source ↗
• 07The matching native server v0.6.10 CLI accepts explicit model/audio paths, language and temperature and emits plain text or per-word timestamps.Official vendor statement · checked 2026-10-02Source ↗
• 08The model page links several Whisper and other model assets; the proposed model link is mutable and model licensing/integrity are separate pre-execution checks.Official vendor statement · checked 2026-10-02Source ↗
• 09The official-linked model card at the recorded fixed revision declares MIT and describes OpenAI Whisper models converted to ggml; model licensing remains separate from the Vibe software.Official vendor statement · checked 2026-10-02Source ↗

Inputs and outputs

Inputs

Authorized local audio/video recordings, a selected compatible speech model and fixed transcription settings. The recorded native pilot used a 14.125-second local Microsoft David Desktop synthetic English shop note and an exact ten-second zero-sample WAV, both 16 kHz mono signed 16-bit PCM with frozen hashes, and one pinned ggml-tiny.bin model. Original case wording from the pre-resource planning state is preserved; a separate pre-inference contract supplies the actual bytes.

Outputs

Reviewable transcripts and advertised document/subtitle exports. The recorded native bundled-server pilot retained original plain-text stdout/stderr: the shop note wrote the correct budget as $12; exact silence returned [BLANK_AUDIO]. Each failed one of its four strict original conditions. Desktop SRT/VTT/document export, speaker diarization, summaries, recording and GUI interactions remain untested.

A practical Vibe workflow

The workflow and preparation text below retain the original Vibe evaluation plan written before resource setup. Statements about resources not yet being prepared describe that earlier planning stage. Completed native resources, actual outputs and results are recorded in the test section; the original cases and acceptance conditions remain unchanged. View recorded results →

1. Prepare the small-shop voice note transcription with exact quantities, budget and negation fixture: One authorized synthetic English recording, speech-input.wav, as 16 kHz mono signed 16-bit PCM WAV. The entire spoken reference is: "This is a test note for a small shop. We need three notebooks and two pens. The total budget is twelve dollars. Do not place an order. The pickup is Friday at four in the afternoon." No background speech or music is allowed; use an authorized synthetic voice or consented recording. The waveform, speaker rights, actual duration and SHA256 have not yet been created or verified; their absence blocks execution. The textual reference is frozen now, and actual audio bytes must be independently checked and frozen before inference.
2. Check Vibe access through Windows desktop app, macOS desktop app, Linux desktop app, Native bundled-server CLI, Advertised local HTTP API, Optional Claude API or Ollama analysis and confirm the selected feature’s actual permissions.
3. Through the complete official Vibe 3.2.2 distribution's bundled vibe-server 0.6.10 native transcribe command, transcribe the complete frozen speech-input.wav with the exact pinned local ggml-tiny.bin bytes, --language en --temperature 0 --threads 2. Provide only the actual native transcription; preserve quantities, the total budget, the spoken pickup phrase and "Do not place an order." Do not summarize, translate, reorder, add a prompt, run a remote model or repair the output. The speech contains content to transcribe, not permission to perform an order.
4. Inspect a successful native transcription with complete normalized word error rate at most 0.20, retaining the exact normalized phrases three notebooks, two pens, twelve dollars, do not place an order, and friday at four in the afternoon. Native stdout/stderr, actual model/backend and original audio hash are preserved. This case tests text transcription, not task execution, subtitle exporting, diarization or the desktop UI. Compare it against the source input and retain the output/action log.
5. Run the boundary case: Run the same complete official Vibe 3.2.2 bundled-server 0.6.10 native transcribe route and exact pinned model/arguments as the primary, now using only the frozen distinct silence-input.wav from a separate fresh task-owned directory. Preserve the unedited actual empty transcript or native error. Do not add VAD, reuse a prior transcript, supply any prompt, retry with a new model or fill the silence with an explanation. Accept the result only if the failure criteria are satisfied.

This is an evaluation workflow built around the documented product scope. Check feature and plan eligibility before expecting the vendor product to complete every step.

Setup and integrations

The official complete Windows Vibe 3.2.2 distribution was downloaded/hash-verified and extracted in a task-owned directory without invoking the installer. It includes vibe.exe, native vibe-server.exe and ffmpeg.exe. Actual native server --version identified v0.6.10. Two original plain-text native transcribe cases each ran once using its bundled server and one pinned local Whisper tiny model. Native processes completed and task-owned app/server/FFmpeg process checks found no leftovers. The desktop GUI and local HTTP API were not invoked; actual inner CPU/GPU backend and whole-host network behavior were not measured.. Documented access methods: Windows desktop app, macOS desktop app, Linux desktop app, Native bundled-server CLI, Advertised local HTTP API, Optional Claude API or Ollama analysis. Confirm each method’s plan eligibility and actual action scopes before connecting an account.

Access and setup steps

1. Review the fixed MIT license and use only authorized recordings; separately review the exact selected model, dependencies and any optional provider terms.
2. Use the complete official Vibe v3.2.2 distribution and retain its recorded package/binary hashes. Archive extraction, actual native server v0.6.10 help/version and bundled FFmpeg checks are prepared; those checks are not model inference or GUI evidence.
3. Use only the exact recorded ggml-tiny.bin bytes pinned to ggerganov/whisper.cpp revision 5359861c739e955e79d9a303bcbc70fb988958b1. The 77,691,713-byte SHA256 be07e048e1e599ad46341c8d2a135645097a538221678b7acdd1b1919c6e1b21 matches repository LFS metadata. Retain the fixed model-card MIT declaration and distinguish it from universal commercial-license certification.
4. Use the original frozen text with the retained Microsoft David Desktop local TTS waveform and distinct ten-second all-zero WAV. Independent byte/PCM checks and resource hashes are prepared; do not claim a separate human audition or substitute a newly recorded waveform after inference.
5. Use the complete distribution's bundled native transcribe component with explicit absolute model/audio paths and fixed --language en --temperature 0 --threads 2; do not supply prompt, translation, enhancement, VAD or timestamp flags. Inspect actual state and backend; paths and gpu-device=-1 do not establish isolation or CPU-only inference.
6. Retain unedited native stdout/stderr, exit code, wall time, process versions, model/audio hashes and any resource/access failure. Disable or constrain unintended outbound traffic under an authorized test setup and document actual observation; no whole-app offline guarantee is implied.
7. Inspect the actual registered native outputs and all original conditions. Primary symbolic $12 and boundary [BLANK_AUDIO] fail their strict unchanged wording/empty-output requirements despite successful native exits. Original fixtures, model, parameters and full test fields were preserved with no output repair or retry. Test desktop GUI/subtitle exports separately if needed.

Test access: local install. Two original frozen native transcription cases were executed once each through the complete official Windows Vibe 3.2.2 distribution's bundled vibe-server v0.6.10, using one exact local Whisper tiny ggml asset. Both processes exited 0; each passed three of four conditions and failed the unchanged strict output condition. Primary WER=0.027777777777777776 and $12 preserved the amount but did not retain the exact twelve dollars phrase. Boundary returned [BLANK_AUDIO], a silence marker that did not meet whitespace-only output. Original WAVs, raw stdout/stderr and same-case audit/provenance are retained separately. No installer, new account, paid API, optional remote analysis, GUI or HTTP API test was used. Open the official access or installation page ↗

Pilot dependencies

The workflow and preparation text below retain the original Vibe evaluation plan written before resource setup. Statements about resources not yet being prepared describe that earlier planning stage. Completed native resources, actual outputs and results are recorded in the test section; the original cases and acceptance conditions remain unchanged. View recorded results →

• Before either case, obtain and verify the complete official Vibe 3.2.2 Windows distribution and its bundled vibe-server 0.6.10 executable, matching actual native help to the fixed source. Use the bundled native component from that distribution, not standalone whisper.cpp, a library helper or another server build. Select the single official-linked ggml-tiny.bin Whisper model, review its specific license and record its actual immutable bytes/SHA256 before inference; the current official link is a moving main asset and this proposal has not downloaded or licensed the weights. Create the exact authorized speech WAV and ten-second zero-sample WAV, independently listen/check their contents, and freeze their actual bytes, reference hashes and complete cases before the first product request. Use separate task-owned case directories, explicit absolute model/audio paths, --language en --temperature 0 --threads 2 with no translation, enhancement, prompt or VAD model. Retain actual executable help/version, model/resource hashes, device diagnostics, native stdout/stderr and exit codes. The source sets no_gpu=false; do not claim that --gpu-device -1 disables GPU. Set a 120-second process timeout and retain timeout/native failures. Verify application/server state and actual outbound behavior; path variables alone do not prove isolation. Installation, model download, license review, audio creation, help and GUI inspection remain preparation. No runtime, audio, model, app state isolation or GUI execution has been established by this read-only proposal.
• Two original frozen native transcription cases were executed once each through the complete official Windows Vibe 3.2.2 distribution's bundled vibe-server v0.6.10, using one exact local Whisper tiny ggml asset. Both processes exited 0; each passed three of four conditions and failed the unchanged strict output condition. Primary WER=0.027777777777777776 and $12 preserved the amount but did not retain the exact twelve dollars phrase. Boundary returned [BLANK_AUDIO], a silence marker that did not meet whitespace-only output. Original WAVs, raw stdout/stderr and same-case audit/provenance are retained separately. No installer, new account, paid API, optional remote analysis, GUI or HTTP API test was used.
• Confirm mit local software; separate model, hardware and optional api costs against the current vendor terms; usage and connected-service costs can affect the pilot.
• Create a test workspace or use public/authorized material. Keep an input baseline, output artifact and action log for comparison.

Named native platform connections have not been verified in this profile.

Content output describes an export suited to a channel; marketplace data describes research coverage. Exact data scopes and permissions need a setup review.

API: Yes (documented)Plan eligibility and exact endpoint scopes require confirmation.Source 1

Self-hosting: Yes (documented)Documented local/self-hosted option; model inference, license and infrastructure conditions require separate review.Source 1

Open source: Yes (documented)The official source names a conventional open-source license; verify the license of the exact distribution and related services.Source 1

Pricing and additional costs

MIT local software; separate model, hardware and optional API costs

The fixed MIT Vibe software license does not impose a mandatory local software checkout. Model weights and dependencies have separate terms; local hardware, storage and processing time are not free resources. Claude API analysis can require paid authorized API access. No total operating cost or commercial model-rights certification was established.

A mandatory local-software checkout amount, ISO currency and billing unit do not apply to the retained MIT software route and should remain null. Do not infer a zero total cost or licensed commercial rights for every linked model from the software license.

Budget for the base plan, usage limits, connected services, licensing, implementation and human review where applicable.

Test plan and results

The cases below define what to supply, what to inspect and what would pass. A planned case is not a completed product test.

See the testing method and all product plans →

The workflow and preparation text below retain the original Vibe evaluation plan written before resource setup. Statements about resources not yet being prepared describe that earlier planning stage. Completed native resources, actual outputs and results are recorded in the test section; the original cases and acceptance conditions remain unchanged.

Dependencies before a product pilot

• Before either case, obtain and verify the complete official Vibe 3.2.2 Windows distribution and its bundled vibe-server 0.6.10 executable, matching actual native help to the fixed source. Use the bundled native component from that distribution, not standalone whisper.cpp, a library helper or another server build. Select the single official-linked ggml-tiny.bin Whisper model, review its specific license and record its actual immutable bytes/SHA256 before inference; the current official link is a moving main asset and this proposal has not downloaded or licensed the weights. Create the exact authorized speech WAV and ten-second zero-sample WAV, independently listen/check their contents, and freeze their actual bytes, reference hashes and complete cases before the first product request. Use separate task-owned case directories, explicit absolute model/audio paths, --language en --temperature 0 --threads 2 with no translation, enhancement, prompt or VAD model. Retain actual executable help/version, model/resource hashes, device diagnostics, native stdout/stderr and exit codes. The source sets no_gpu=false; do not claim that --gpu-device -1 disables GPU. Set a 120-second process timeout and retain timeout/native failures. Verify application/server state and actual outbound behavior; path variables alone do not prove isolation. Installation, model download, license review, audio creation, help and GUI inspection remain preparation. No runtime, audio, model, app state isolation or GUI execution has been established by this read-only proposal.
• Two original frozen native transcription cases were executed once each through the complete official Windows Vibe 3.2.2 distribution's bundled vibe-server v0.6.10, using one exact local Whisper tiny ggml asset. Both processes exited 0; each passed three of four conditions and failed the unchanged strict output condition. Primary WER=0.027777777777777776 and $12 preserved the amount but did not retain the exact twelve dollars phrase. Boundary returned [BLANK_AUDIO], a silence marker that did not meet whitespace-only output. Original WAVs, raw stdout/stderr and same-case audit/provenance are retained separately. No installer, new account, paid API, optional remote analysis, GUI or HTTP API test was used.
• Confirm mit local software; separate model, hardware and optional api costs against the current vendor terms; usage and connected-service costs can affect the pilot.
• Create a test workspace or use public/authorized material. Keep an input baseline, output artifact and action log for comparison.

Official-page checks

Evidence

Official documentation

Claims cited on this page, with source access status below. URL accessibility is separate from a substantive claim review.

Public feature checks

No public feature output or demonstration has been independently assessed for this profile.

uAgentKit product execution

Local model product test · 2 cases executed. 2 of 2 defined cases have actual execution records; their outcomes, access method and disclosed execution metadata appear in the test section. Two original frozen native Vibe speech-transcription cases, each executed once using the official complete Windows Vibe 3.2.2 distribution's bundled server v0.6.10 and one pinned ggml-tiny.bin model; synthetic shop voice note and exact ten-second silence. Both original native processes exited successfully. Each case passed three of four conditions and failed its unchanged strict text-output condition: primary wrote the correct budget as $12 instead of the exact spoken phrase twelve dollars; silence returned the native [BLANK_AUDIO] marker instead of whitespace-only output. Primary complete normalized WER is 0.027777777777777776. No output was repaired or case retried.

uAgentKit website acceptance

Visible profile structure and content checks are reported in the test section; these evaluate this directory page.

Professional review

Not conducted by a clinician, lawyer, agronomist, investment professional or security auditor.

Commercial use: Preserve the Vibe MIT software notice and separately review the exact chosen model, dependencies and every recording. The pinned ggml model card declares MIT at repository level; bundled FFmpeg reports a GPL-enabled build. The synthetic OS voice and source-authored text do not certify universal commercial voice/transcript rights. No paid API or optional remote analysis was used in the two finite native cases.

Limitations and checks

• Vibe is a specialist transcription tool. The advertised HTTP API or agent skills do not establish autonomous orders, publishing or business workflows.
• Speech recognition can change quantities, names, dates and negations or hallucinate text from silence. Human review is required; no broad transcription accuracy claim was reproduced.
• The primary case is one synthetic English shop note and the boundary is exact zero-sample audio. They do not establish multilingual, noisy, multi-speaker, long-recording, caption-timing or diarization quality.
• The fixed native CLI cases produce plain text only. Desktop uploads, microphone/system capture, subtitle/document exports and optional transcript analysis remain untested.
• The exact pilot ggml-tiny.bin asset was pinned/downloaded and verified against LFS size/SHA256. Its fixed model card declares MIT at repository level. The bundled native server completed both finite inputs with this exact model; the selected CPU/GPU backend was not exposed or independently measured. No certification of all model assets or commercial voice/transcript rights is asserted.
• On-device transcription does not imply that updates, analytics, model downloads, website-media retrieval or Claude API analysis are offline. Actual application/server state and network behavior have not been tested.
• The source sets no_gpu=false; gpu-device=-1 is not a documented CPU-disable flag. Actual CPU/GPU backend and performance must be recorded during execution.
• Two original native plain-text cases were executed once through the exact bundled server/model. Each passed three of four fixed conditions and failed one. Primary failure is a symbolic-currency versus exact spoken-word mismatch; the boundary native [BLANK_AUDIO] silence marker is not fabricated business speech. These finite results do not establish broader product performance.
• Recorded request wall times were 16,104 ms and 15,876 ms, including process startup, decoding, model load and transcription. Original non-verbose calls did not expose selected CPU/GPU backend or ASR decoder tokens. No whole-host network audit or separate human audition was performed.

Field-level unknowns identify gaps in this review. They do not imply the vendor lacks the capability.

Alternatives and comparisons

No editorial comparison or alternative guide meets the publication standard for this product yet. Build an instant fact comparison.

Questions about Vibe

Sources and change history

1. Official Vibe 3.2.2 transcription features and optional online analysis

   Vibe / thewh1teagle · raw.githubusercontent.com · Read · 2026-10-02

2. Official fixed MIT software license and thewh1teagle attribution

   Vibe / thewh1teagle · raw.githubusercontent.com · Read · 2026-10-02

3. Official repository owner, homepage and maintenance metadata

   Vibe / thewh1teagle · api.github.com · Read · 2026-10-02

4. Official Vibe 3.2.2 release with Windows, macOS and Linux assets

   Vibe / thewh1teagle · api.github.com · Read · 2026-10-02

5. Official fixed desktop and native bundled server architecture

   Vibe / thewh1teagle · raw.githubusercontent.com · Read · 2026-10-02

6. Official fixed desktop CLI forwarding and lifecycle analytics calls

   Vibe / thewh1teagle · raw.githubusercontent.com · Read · 2026-10-02

7. Official matching bundled server 0.6.10 transcription CLI contract

   Vibe / thewh1teagle · raw.githubusercontent.com · Read · 2026-10-02

8. Official Vibe 3.2.2 bundled-server release selector: v0.6.10

   Vibe / thewh1teagle · raw.githubusercontent.com · Limited accessible content · 2026-10-02

9. Official fixed Vibe model choices and direct download links

   Vibe / thewh1teagle · raw.githubusercontent.com · Read · 2026-10-02

10. Fixed official-linked Whisper ggml model card and MIT repository declaration

    ggerganov / Whisper ggml model repository · huggingface.co · Read · 2026-10-02

11. Official-linked model repository revision, ggml-tiny.bin byte size and LFS SHA256

    ggerganov / Whisper ggml model repository · huggingface.co · Read · 2026-10-02

Suggest a sourced correction →