68 questions about video & audio.

Video Compressor is free for one lifetime use after sign-up — no credit card required. For unlimited use, the Basic plan (£7.99/mo) unlocks it. The Free tier exists so you can try every tool once with no payment commitment.

Video Compressor accepts up to 100 MB on Free, 500 MB on Basic / Pro, and 2 GB on Plus+. Hard cap is 25 MB on Whisper-based tools (OpenAI API limit, not ours).

Video Compressor uploads your file over HTTPS to our EU-resident infrastructure (Cloudflare R2). The file is held while processing then auto-deletes after 24 hours (Free) or 7 days (paid). Never sold, never used to train any AI model.

Yes — Video Compressor works on modern mobile browsers (iOS Safari 17+, Android Chrome 120+). Heavy in-browser tools (large PDFs, video) work best on desktop. A dedicated PWA ships in V2 month 4–6 for a more app-like mobile experience.

No — Video Compressor never adds a watermark, on Free or any paid plan. The Free tier limits use to 1 per tool to fund the platform, not by watermarking.

Yes — the public API at /api/tools/* is available from Pro tier and above (£17.99/mo). 10,000 calls per month on Pro, 50k on Plus, 200k on Power, 1M on Agency. Full reference at /docs/api (V2 launch).

Video Compressor works on the Free tier (1 use) and any paid plan unlimited. Each paid tier raises the monthly cap on the meter Video Compressor uses.

Batch mode in the UI ships in V2 — for now Video Compressor handles one file at a time. Pro+ subscribers with API access can build their own batch loop via the public API (10k+ calls/month available).

On the Hero 10 tools (incl. Chat with PDF, AI Resume Tailor, Background Remover, AI Headshot, Photo Restoration, Auto-Subtitle, AI Slide Deck, Image Upscaler, PDF Summarizer, AI Summarizer) — yes, we invest 60% of polish budget there. Video Compressor is built to be 95%+ as good as the specialist tool you'd pay $20-79/mo for. On the long-tail 90 tools, Video Compressor is 70-80% of specialist quality — sufficient for casual use.

Auto Subtitle Generator is free for one lifetime use after sign-up — no credit card required. For unlimited use, the Basic plan (£7.99/mo) unlocks it. The Free tier exists so you can try every tool once with no payment commitment.

Auto Subtitle Generator accepts up to 100 MB on Free, 500 MB on Basic / Pro, and 2 GB on Plus+. Hard cap is 25 MB on Whisper-based tools (OpenAI API limit, not ours).

Auto Subtitle Generator uploads your file over HTTPS to our EU-resident infrastructure (Cloudflare R2). The file is held while processing then auto-deletes after 24 hours (Free) or 7 days (paid). Never sold, never used to train any AI model.

Yes — Auto Subtitle Generator works on modern mobile browsers (iOS Safari 17+, Android Chrome 120+). Heavy in-browser tools (large PDFs, video) work best on desktop. A dedicated PWA ships in V2 month 4–6 for a more app-like mobile experience.

No — Auto Subtitle Generator never adds a watermark, on Free or any paid plan. The Free tier limits use to 1 per tool to fund the platform, not by watermarking.

Yes — the public API at /api/tools/* is available from Pro tier and above (£17.99/mo). 10,000 calls per month on Pro, 50k on Plus, 200k on Power, 1M on Agency. Full reference at /docs/api (V2 launch).

Auto Subtitle Generator works on the Free tier (1 use) and any paid plan unlimited. Each paid tier raises the monthly cap on the meter Auto Subtitle Generator uses.

Batch mode in the UI ships in V2 — for now Auto Subtitle Generator handles one file at a time. Pro+ subscribers with API access can build their own batch loop via the public API (10k+ calls/month available).

On the Hero 10 tools (incl. Chat with PDF, AI Resume Tailor, Background Remover, AI Headshot, Photo Restoration, Auto-Subtitle, AI Slide Deck, Image Upscaler, PDF Summarizer, AI Summarizer) — yes, we invest 60% of polish budget there. Auto Subtitle Generator is built to be 95%+ as good as the specialist tool you'd pay $20-79/mo for. On the long-tail 90 tools, Auto Subtitle Generator is 70-80% of specialist quality — sufficient for casual use.

Audio Transcription is free for one lifetime use after sign-up — no credit card required. For unlimited use, the Basic plan (£7.99/mo) unlocks it. The Free tier exists so you can try every tool once with no payment commitment.

Audio Transcription caps audio uploads at 25 MB (Whisper API hard limit). For longer recordings, split first or transcribe locally.

Audio Transcription uploads your file over HTTPS to our EU-resident infrastructure (Cloudflare R2). The file is held while processing then auto-deletes after 24 hours (Free) or 7 days (paid). Never sold, never used to train any AI model.

Yes — Audio Transcription works on modern mobile browsers (iOS Safari 17+, Android Chrome 120+). Heavy in-browser tools (large PDFs, video) work best on desktop. A dedicated PWA ships in V2 month 4–6 for a more app-like mobile experience.

No — Audio Transcription never adds a watermark, on Free or any paid plan. The Free tier limits use to 1 per tool to fund the platform, not by watermarking.

Yes — the public API at /api/tools/* is available from Pro tier and above (£17.99/mo). 10,000 calls per month on Pro, 50k on Plus, 200k on Power, 1M on Agency. Full reference at /docs/api (V2 launch).

Audio Transcription works on the Free tier (1 use) and any paid plan unlimited. Each paid tier raises the monthly cap on the meter Audio Transcription uses.

Batch mode in the UI ships in V2 — for now Audio Transcription handles one file at a time. Pro+ subscribers with API access can build their own batch loop via the public API (10k+ calls/month available).

On the Hero 10 tools (incl. Chat with PDF, AI Resume Tailor, Background Remover, AI Headshot, Photo Restoration, Auto-Subtitle, AI Slide Deck, Image Upscaler, PDF Summarizer, AI Summarizer) — yes, we invest 60% of polish budget there. Audio Transcription is built to be 95%+ as good as the specialist tool you'd pay $20-79/mo for. On the long-tail 90 tools, Audio Transcription is 70-80% of specialist quality — sufficient for casual use.

OpenAI Whisper large-v3 — word-level accuracy ~95-98% on clear English, ~92-95% on accented English, ~85-92% with heavy background music. Speaker labels on Pro+ via Pyannote.

OpenAI's Whisper API has a 25 MB hard cap. For longer recordings, split first (any free audio editor) or compress to ~64 kbps MP3 first. We're testing a chunking pipeline for V2 to remove the limit.

33 dropdown options + auto-detect. Whisper internally supports 100+ languages; the 33 in our UI are the highest-quality / most-requested. Email hello@apexkit.pro to enable others.

SRT is the older format — works with YouTube, Vimeo, video editors. WebVTT is the modern web standard — works with HTML5 video, modern players. Subtitle Generator exports both.

Burn-in (so the subtitles can't be hidden) ships in V2 alongside the broader video-editing tools. For now we export SRT/WebVTT sidecar — TikTok and Instagram both auto-burn from upload.

Not in subtitle output (uncommon in standard SRT). Use Transcribe Audio for speaker-labelled output, then convert to SRT manually if needed.

High (~70% of source size, near-lossless), Balanced (~40% of source, still 1080p-clean), Small (~20% of source, best for social). All output H.264 + AAC in fragmented MP4 for streaming-friendly upload.

No — output resolution matches input. Quality is controlled via CRF (constant rate factor); lower CRF = higher quality.

MP4, MOV, WebM, MKV, AVI — anything ffmpeg can ingest. Output is always MP4 (H.264 + AAC) for broad compatibility.

Video trim ships V2 (need a small ffmpeg wrapper). For now, most desktop video apps (QuickTime, Photos, OBS) handle trim, then Apexkit compresses the result.

Video → GIF ships V2. For now: extract a short clip in a desktop tool, then upload to a free GIF converter; we'll add native support soon.

Same Whisper backbone under the hood. Otter's edge is real-time meeting capture + calendar integration. Apexkit's edge is bundled pricing + async transcription quality. See /compare/otter-ai.

Yes — Transcribe Audio accepts video files; it strips the audio internally before transcribing. To save the audio as a standalone MP3, V2 ships a dedicated 'extract audio' tool.

Voice cloning ships V2 month 4-6 — via ElevenLabs API, with consent verification. Not currently available.

TTS ships V2 month 4-6. Will use OpenAI TTS or ElevenLabs for the high-quality voices.

Audio denoise ships V2 — likely via Replicate's RNNoise model. Today the closest workaround: a desktop tool like Audacity (noise reduction is one click).

Podcast clip generator ships V2 — Whisper transcript + GPT-4o ranking + ffmpeg clipping. Targeted for content creators.

Subtitle translation ships V2 — Whisper transcribes, then GPT-4o translates each cue. Today: transcribe → copy plain text → use the Translate tool → manually re-cue.

Two-step today: Transcribe Audio → paste output into AI Summarizer. Native 'meeting summary' (action items + key decisions) ships V2.

MP3 is faster (just audio data) — upload it if you have a choice. MP4 video works too; we extract the audio internally before sending to Whisper.

Yes — Zoom exports as MP4 or M4A, both supported. Set Zoom to record audio only (M4A) for faster Apexkit uploads.

Yes — Teams exports as MP4 (downloaded from Stream / OneDrive). Upload directly to Transcribe Audio or Subtitle Generator.

Yes — Meet saves to Drive as MP4. Download from Drive, upload to Apexkit.

Whisper handles vocals over music reasonably (85-92% accuracy depending on mix). For accurate lyrics, use a specialised music-transcription tool; we're not optimised for that.

Use Transcribe Audio (outputs plain text). Subtitle Generator outputs timestamped formats (SRT/VTT). They're cousins — use the right one for your output need.

Audio merge ships V2 — straightforward ffmpeg job. Until then, use a desktop tool (Audacity, GarageBand) for the merge then upload to Apexkit.

AAC at 128 kbps — sounds transparent on speech, near-transparent on most music. Lossless audio (FLAC inside MKV) ships V2.

Video rotate ships V2 alongside trim + crop. ffmpeg handles it; the UI wrapper is on the roadmap.

Watermark overlay ships V2 month 4-6. Today closest is FFmpeg command-line (out of scope for us).

MP4 only today (H.264 + AAC) for max compatibility. WebM / AV1 output ships V2 — they're smaller but less universally supported.

Input yes (ffmpeg handles it). Output is H.264 (best compatibility); H.265 output ships V2 — saves ~30% file size but smaller player support.

Not by default — output resolution matches input. We're adding a resolution preset (1080p, 720p, 480p) in V2 for explicit downscaling.

~3-4 minutes on the standard queue; ~90 seconds on Pro+ priority queue. Total cost per 1-hour transcription: ~$0.36 in API spend — bundled into your Pro tier.

Whisper has a translate-to-English mode (Spanish audio → English text). Our UI exposes it via a 'translate output to English' toggle (V2). For now: transcribe first, then run output through the Translate tool.

Aspect ratio change with smart-crop ships V2 — targeted at TikTok / Reels reformat from YouTube clips. ffmpeg handles the underlying crop.

Compress Video downmixes to stereo (best mobile compatibility). Multi-track preservation ships V2 if there's demand.

Audio replacement / overlay ships V2. ffmpeg handles it; UI in development.

Video transitions ship V2 alongside trim + crop. ffmpeg fade filters are simple to expose.

Video loop tool ships V2. Today: most video editors can do this in one step (DaVinci Resolve, iMovie).

Slideshow video generator ships V2. For now, AI Slide Deck creates a PPTX that PowerPoint / Keynote can export as MP4.

Frame extract ships V2. Today: any browser-based tool, or pause + screenshot.