🔒 Verified source: github.com/SYSTRAN⏳ Pending test
📄🧭🔗📚🧭🗺️🔗
🧩 Teach this AI new skills
Unlock 8 new powers for your AI agent
faster-whisper has no skill store we can verify. 7 popular cross-AI skills for agents like Claude Code and Cursor, mapped to what your AI agent can do after you install them · tap a power
Pre-check: open Task Manager → Performance to see RAM and GPU. Then run this in Command Prompt to see GPU name, driver version and VRAM. Also check free disk space.
nvidia-smi
Do you see your GPU name and "Driver Version"?
Install Python 3.9+, then faster-whisper in a venv.
Create transcribe.py with this code (replace audio.mp3 with your file). No separate ffmpeg needed.
from faster_whisper import WhisperModel
model = WhisperModel("small", device="cpu", compute_type="int8")
segments, info = model.transcribe("audio.mp3", language="hi")
for s in segments:
print(f"[{s.start:.1f}s -> {s.end:.1f}s] {s.text}")
Run it. The first run downloads the model (~500 MB).
python transcribe.py
Text with timestamps printed?
For NVIDIA GPU use device="cuda", compute_type="float16". This needs cuBLAS for CUDA 12 + cuDNN 9 — see the README GPU section.
Create transcribe.py with this code (replace audio.mp3 with your file). No separate ffmpeg needed.
from faster_whisper import WhisperModel
model = WhisperModel("small", device="cpu", compute_type="int8")
segments, info = model.transcribe("audio.mp3", language="hi")
for s in segments:
print(f"[{s.start:.1f}s -> {s.end:.1f}s] {s.text}")
Run it. The first run downloads the model (~500 MB).
python transcribe.py
Text with timestamps printed?
Pre-check: Apple menu → About This Mac. Note the Chip (Apple M-series or Intel), Memory (RAM) and macOS version.
Noted chip, memory and macOS version?
Install with Python 3.9+ (Macs use CPU int8 mode).
Create transcribe.py with this code (replace audio.mp3 with your file). No separate ffmpeg needed.
from faster_whisper import WhisperModel
model = WhisperModel("small", device="cpu", compute_type="int8")
segments, info = model.transcribe("audio.mp3", language="hi")
for s in segments:
print(f"[{s.start:.1f}s -> {s.end:.1f}s] {s.text}")
Run it. The first run downloads the model (~500 MB).
python transcribe.py
Text with timestamps printed?
Pre-check: check RAM, disk and GPU in a terminal.
free -h
df -h ~
lspci | grep -i -E "vga|3d"
See RAM, free disk and GPU name?
Install. For NVIDIA GPUs also add the pip packages from the README.
Create transcribe.py with this code (replace audio.mp3 with your file). No separate ffmpeg needed.
from faster_whisper import WhisperModel
model = WhisperModel("small", device="cpu", compute_type="int8")
segments, info = model.transcribe("audio.mp3", language="hi")
for s in segments:
print(f"[{s.start:.1f}s -> {s.end:.1f}s] {s.text}")
Run it. The first run downloads the model (~500 MB).
python transcribe.py
Text with timestamps printed?
Pre-check: see your RAM (Task Manager / About This Mac / free -h). Small tools run without a GPU — just slower.
Know how much RAM you have?
Same as your OS tab — this already uses CPU int8. On slow PCs replace "small" with "base".
Run it. The first run downloads the model (~500 MB).
python transcribe.py
Text with timestamps printed?
Sizes are approximate and can change on the official page.
Errors & fixes
"Could not load library cudnn_ops64_9.dll" / libcudnn error
cuDNN 9 missing in GPU mode. Install cuBLAS/cuDNN per README, or run with device="cpu", compute_type="int8".
ValueError: Requested float16 compute type…
float16 does not work on CPU — use compute_type="int8".
"python is not recognized" / "command not found: python"
Python is missing or not on PATH. Install from python.org and tick "Add python.exe to PATH" on the first installer screen, then open a new terminal. On Mac/Linux use python3.
Download stuck / very slow / "connection reset"
Network or DNS issue. Restart the router, try a mobile hotspot, or set DNS to 1.1.1.1 / 8.8.8.8. Re-running the same command usually resumes the download.
Popular AI skills that work with tools like faster-whisper 7 skills
faster-whisper has no skill store we can verify. These cross-AI skills (MCP servers and Agent Skills) work in most AI agents — each card shows one example install; see all AIs for yours.
MCPPlugins
🛡 Stay safe
Read first
Official first — from the AI maker or the service’s own GitHub org / domain.
Read SKILL.md and any scripts before installing. Skills can run code with your permissions.
Never paste base64 or “curl … | sh” download-and-run lines from a skill.
Stars, license and last update come from GitHub (17 Nov 2026). Commands are copied from official docs and are not tested on a real machine yet. Access tags are our summary — always check the source. Skills for every AI →