Specs

System requirements

StreamStash runs on most Windows PCs from the last six or seven years. Heavier features (24/7 monitoring, AI semantic search, hardware-accelerated compression) reward better hardware, but nothing here is exotic.

Minimum

Free · Personal

Casual use, a handful of feeds, the occasional live capture.

  • OSWindows 10 or 11, 64-bit
  • CPU4-core, ~2018 or newer (Intel i3 8th gen / Ryzen 3 2000 series and up)
  • RAM8 GB
  • GPUAny. Falls back to CPU encoding (libx264) if no hardware encoder is detected.
  • Storage10 GB free for the app and cache, plus whatever you allocate for recordings.
  • BrowserFirefox or Librewolf for cookie access (see notes below).
  • RuntimeMicrosoft WebView2 (installed automatically if missing).
  • NetworkStable broadband. Wired recommended for long live captures.

Recommended

Power

24/7 monitoring across many platforms, AI semantic search, cross-platform dedup.

  • OSWindows 10 or 11, 64-bit
  • CPU6 to 8 cores (Ryzen 5 5000 series / Intel i5 12th gen or newer)
  • RAM16 GB (the CLIP semantic-search model and ffmpeg both sit resident in memory)
  • GPUNVIDIA with NVENC (GTX 1660 / RTX 2060 or newer). Intel QSV and AMD AMF also work but get less testing.
  • Storage1 TB+ dedicated drive for recordings. A separate disk is ideal; the recordings folder can be relocated at any time.
  • BrowserFirefox or Librewolf.
  • NetworkStable broadband, wired.

Things worth knowing before you install

Why Firefox for cookies

StreamStash never asks for platform passwords. It reads your existing browser session cookies for the platforms you want to monitor. Firefox stores cookies as a plain SQLite file the app can read in milliseconds. Chrome, Edge, and Brave (since v127) encrypt cookies with Windows DPAPI and keep an exclusive lock while the browser or any background process is running, so the app reads zero cookies even though everything looks healthy. Use Firefox or Librewolf and the friction disappears.

GPU encoders, in priority order

StreamStash auto-detects what your machine can do and cascades down: NVIDIA NVENC first, then Intel QuickSync (QSV), then AMD AMF, finally CPU-only libx264 if nothing else is available. The CPU fallback works on any machine but uses noticeably more energy and time per recording. NVENC is the smoothest path, which is why it's our top recommendation.

RAM and the AI search model

Power-tier AI semantic search loads a CLIP model into memory the first time it runs and keeps it warm. On 8 GB systems this is workable but tight; 16 GB makes it comfortable, and you'll feel the difference if you're running a browser, OBS, or anything else at the same time. AI search is a Power-only feature, so this only matters if you're on Power.

Storage planning

The app itself is small. Recording archives are not. A few monitored TikTok creators can produce hundreds of GB over a year; multi-platform archiving scales accordingly. We recommend pointing the recordings folder at a dedicated HDD or SSD (the in-app Storage settings let you change the location at any point without re-downloading anything).

macOS, Linux, Docker

Windows is our current platform. macOS, native Linux, and Docker images (Synology, QNAP, Unraid) are on the roadmap in that order. If you buy a paid licence now, the upgrade path to those platforms is included; there is no separate charge.

Older or low-end machines

StreamStash will run on hardware below the minimum spec; we just can't promise the same experience. Live recording in particular is bandwidth- and CPU-sensitive: an older laptop with no hardware encoder can still capture a TikTok Live, but post-processing will lean on the CPU and take longer. Try the free tier first to see how your machine handles it.