whisper.cpp
Turn speech into text on your own computer

Images from the project's own README.
What it does
whisper.cpp transcribes audio recordings into text using OpenAI's Whisper speech model, running entirely on your computer. It is a command-line tool, so it is used by typing commands.
Who it's for
Technical users who need private, offline transcription of interviews, lectures or meetings.
Key features
- Offline speech-to-text
- Supports many languages
- Runs efficiently on laptops
Often used instead of: Otter.ai, Rev
Compatibility
NOVI never presents a guess as verified.- WindowsUnknown
No official build, package or documentation for this system was found.
- macOS (Apple Silicon)Unknown
No official build, package or documentation for this system was found.
- macOS (Intel)Unknown
No official build, package or documentation for this system was found.
- LinuxUnknown
No official build, package or documentation for this system was found.
What do these labels mean?
- Confirmed
- NOVI users have completed verified installations on this system.
- Evidence-supported
- The project publishes an official build or package for this system.
- Inferred
- Likely to work based on how the project is built, but not confirmed.
- Unknown
- We have not found enough evidence either way.
- Unsupported
- The project does not support this system.
How to get it
Guided installationNOVI walks you through the official installation steps, one at a time. Difficulty: Advanced.
Follow the project's own instructions
NOVI hasn't reviewed a download or package for this project yet. The project explains how to install it on GitHub; it may require technical knowledge.NOVI never installs software from your browser. Downloads come directly from the project or its publisher. Only install software you trust.
Get it working
A free, personal setup plan for your computer: the official download, exact steps, and a checklist. Saved in your dashboard, with help if you get stuck. NOVI never installs or runs anything on your computer.
Trust and security
- license
Published under the MIT License (MIT).
Licence detected by GitHub from the repository's licence file. Read the full licence before modifying or redistributing.
Source
NOVI checks licences, maintenance activity and distribution sources. NOVI has not audited this project's code and does not certify it as secure. Only install software you trust.
Reviews
No reviews yet
Be the first to share your experience with whisper.cpp. NOVI never shows invented reviews.
Sign in to write a review.
Recent releases
From GitHubv1.9.5Oct 6, 2026
## Overview Maintenance release. Updates to [ggml v0.26.0](https://github.com/ggml-org/ggml/releases/tag/v0.26.0). **Nightly build:** [b5454](https://github.com/ggml-org/whisper.cpp/releases/tag/b5454) **More info:** [dist : releases and versioning of ggml-org projects](https://github.com/ggml-org/ggml/discussions/1579) ## Changelog since v1.9.4 d1be6fde whisper : bump version to 1.9.5 (#4102) 4afec37b talk-llama : update llama.cpp to v0.6.0 3d451abe sync : ggml 47f1c7dc ggml : bump version to 0.26.0 (ggml/1652) e71b7844 CUDA: make the alloc_deps check batch independent (llama/29986) cdc87302 vulkan: fix Flash Attention shmem write out of bounds (llama/29988) 563a9f7a vulkan: revert mul_mat_id tile selection PR #29182 (llama/29936) ebb8a170 cuda: use the vector lightning indexer kernel on MUSA (llama/29990) 8c6ca4ab CUDA: Optimize accumulation in mmq for NVFP4 type (llama/29857) 82fa9df7 vulkan: fix stale prealloc_y reuse across flash attention and soft_max (llama/29591) f6915422 vulkan: sparse flash attention for quantized K/V (llama/29639) d477de3c llama : fix unexpected graph reallocation in the k-pool models (llama/29958) 95283ca7 ggml-cpu : add Q8_0 IME1 matrix kernel for
Full release notesb5454Pre-releaseOct 6, 2026
No release notes were published.
Full release notesv1.9.4Sep 11, 2026
## Overview New version has been released. **Nightly build:** [b5130](https://github.com/ggml-org/whisper.cpp/releases/tag/b5130) **More info:** [dist : releases and versioning of ggml-org projects](https://github.com/ggml-org/ggml/discussions/1579) ## Changelog since v1.9.3 927cfce3 metal : remove leftover ggml-metal.metal kernels file (#4051) dd80eb2a scripts : use sed instead of grep for version parsing [no ci] (#4052) 1fa6dfa0 ci : update WoA CUDA 13.4 to use 13.4.1 GA (#4053) a2b36eb6 whisper : bump version to 1.9.4 (#4050) 6fb4cd67 ci : add Windows On ARM support to release job (#4048) c44b60b8 whisper : call encoder_begin_callback before language auto-detect (#3936) 6d0ed914 whisper : re-seed decoder 0 between calls (#4025) cec4dbe5 ci : update close-issue job to not close issues (#4045) 79f2d921 tests : load backends before init when built with GGML_BACKEND_DL (#4031) 61e6ccad server : return language in detect response (#4035) 52a939a2 sync : ggml a937f4e8 ggml : bump version to 0.23.0 (ggml/1618) 11d4eec8 metal : add remaining fa-vec tunings for M3 Max (llama/28373) 140e57a4 ggml : replace compile definitions with version.h.in (llama/28364) e2389eb9 ggml : rename and
Full release notesb5130Pre-releaseSep 11, 2026
No release notes were published.
Full release notesb5127Pre-releaseSep 10, 2026
**Note:** the Windows arm64 (CUDA) build (`whisper-bin-win-cuda-*-arm64.zip`) uses a preview edition of the CUDA Toolkit for Windows on Arm and should be considered experimental.
Full release notes
Where this information comes from
- Description, audience and features: Written by NOVI editors from the project's README and official website. link
- Statistics, licence and releases: GitHub API link
GitHub data last refreshed 3 days ago. whisper.cpp is the property of its authors. NOVI is not affiliated with this project.
