Umbiko wephutha / Umbuzo wezici
Umkhiqizi wevidiyo we-AI Umhleli wevidiyo we-AI Umthuthukisi wevidiyo we-AI I-AI Video Dubbing Studio Ukuthatha ukusebenza kwe-AI (Indlela yokushayela-Ingxenye yesi-2) Umhumushi wevidiyo we-AI effects-style Umlinganisi wevidiyo we-AI Okuningi →

Isiqinisi sevidiyo se-AI

Ukusetshenziswa kwebhizinisi OK Amamodeli angama-380+ Akukho phawu lwamanzi Akukho ubhaliso olungenayo
Imodeli:
+ GPT-5, Claude, Gemini
Upload a shaky clip — we run a two-pass ffmpeg vid.stab workflow (optical-flow detect + rewarp) and return a gimbal-smooth MP4. Best results on handheld walking footage, drone clips, and bodycams. Heavy job (~2–3× realtime) so cost scales with duration.

Thola ividiyo engqubuzana lapha noma chofoza ukulayisha

MP4, MOV, WebM kuze kube yi-200MB — ama-clip ancane aphele ngokushesha

— — — —
Subtle ~1% crop, Medium ~2%, Maximum ~4%. Enamandla = crop engaphezu kodwa smoother output.
Ukuzongeza-ukufaka gcina isixazululo se output sifana ne-input.
Isilinganiso se-token se-clip yakho
—
Thenga ama-token
Layisha phezulu ividiyo ukuze ubone izindleko ezifanele zevidiyo yakho.
Layisha phezulu

Uhlobo lwevidiyo oluxazululwe yilesi

Ama-clip ahambayo aphathekayo

Vlogs, ukuvakasha B-roll, izinyawo ezihambayo - vid.stab ilungisa i-rolling footfall bounce ngaphandle kwefreyimu.

Izithombe ze-drone ne-gimbal

Ihlanza ukushaya kwe-prop kanye nokushaya okuncane okuholela emkhathini. Ukuqina okuphakathi kuyindlela engcono kakhulu.

I-bodycam ne-action-cam

GoPro, Insta360, Osmo Action - sebenzisa ngempumelelo amandla amakhulu wokugaya ongekho emthethweni.

Indlela i-vid.stab isebenza ngayo

Two passes. Pass one runs vidstabdetect — optical-flow analysis that tracks hundreds of visual features per frame and writes out a .trf motion file. Pass two runs vidstabtransform, which uses that motion file to build a smooth virtual camera path and rewarps each frame to match it. The result is what a physical gimbal would have given you on the original shoot.

vs. DaVinci Resolve / Premiere / iMovie

Resolve ine-stabilizer enhle, kodwa ingaphansi kwe-1.2GB yokufakela kanye ne-project timeline. Premiere ibiza i-$23/mo. I-Smooth Clip ye-iMovie ilula kodwa i-naive. vid.stab — okuyilokho i-ffmpeg esebenzisa ngaphansi kwe-hood lapha — iyi-library efanayo OpenShot ne-Avidemux ship. Ukuqinisa oku-off, ukufaka i-browser 10 × okulula kunalokho ukuvula iphrojekthi.

Uma UNGAKUFUNDA lokhu

  • Isithombe esithweswe ikhamera esisetshenziswa ngokuvamile (DJI RS, Ronin) — sikhona; isiyingi sesibili silungisa futhi sifaka ukungqubuzana.
  • Ukuzingela ikhamera (i-whip pans, i-dolly shots) — i-vid.stab izobhekana nazo futhi isuse ukunyakaza okulindelekile.
  • I-jello ye-shutter ehambayo enamandla (ifoni ebiza kakhulu ngejubane eliphakeme le-shutter) — i-vid.stab ayikwazi ukususa i-CMOS artifacts. Yithwebula kabusha nge-shutter angle ephansi uma kungenzeka.
Izinketho ezithuthukisiwe
Imiphumela
Ama-token aphansi. Thola ama-token amaningi
Ufuna izimpendulo ezinhle? Amamodeli aphezulu (GPT-5, Claude, Gemini) inikeza umgangatho ophezulu. Bona ama-plans

❤️ Uthanda i-Free.ai? Ngisho nabahlobo bakho!

Ngena ukuze uthole i-link yokuxhumanisa futhi uthole ama-token angama-25,000 ngamunye umngane.

Ufuna okuningi? Bhala mahhala nge-30K tokens / ngosuku + 10K ibhonasi
Bhala

Kuqedwa isicelo sakho...

Ukwenza amavidiyo asheshayo asebenze nge-AI emahhala. Ukwenza amavidiyo abukekayo, asebenza kahle ngokushesha.

Indlela Yokusetshenziswa Isiqinisi sevidiyo se-AI

1
Faka umngenisi wakho

Bhala umbhalo, ulayishe ifayela, noma uchaza ofuna ukukwenza. Akukho akhawunti edingekayo.

2
Chofoza ukwakha

I-AI yethu isebenza isicelo sakho emaminithini usebenzisa amamodeli avulekile angcono kakhulu.

3
Layisha phezulu futhi uhlukanise

Layisha phezulu, kopela, noma uhlukanise imiphumela yakho. Imahhala ukusetshenziswa komuntu siqu nokwebhizinisi.

Sebenzisa lesi sisetshenziswa nge-API

Yenza lokhu sixhobo sisebenze ngokuzenzakalela kusuka kukhodi yakho. I-REST ehambisanayo ne-OpenAI, ukuqinisekisa i-token ye-bearer, akukho SDK edingekayo. Izindleko ze-token zifana ne-web interface.

Ukufaka incwadi Thola isithonjana se-API 
curl -X POST https://api.free.ai/v1/video/generate/ \
  -H "Authorization: Bearer sk-free-..." \
  -H "Content-Type: application/json" \
  -d '{"prompt": "A cat playing piano", "duration": 4}'

Amathuluzi ahlobene ne-AI emahhala

Umkhiqizi wevidiyo we-AI
Umhleli wevidiyo we-AI
Umthuthukisi wevidiyo we-AI
I-AI Video Dubbing Studio
Ukuthatha ukusebenza kwe-AI (Indlela yokushayela-Ingxenye yesi-2)
Umhumushi wevidiyo we-AI
effects-style
Umlinganisi wevidiyo we-AI

Isiqinisi sevidiyo se-AI — FAQ

Layisha phezulu i-clip ephathekayo engqubuzana futhi siyihlanze usebenzisa i-ffmpeg vid.stab — i-optical-flow stabilizer ehamba kabili ethola umthwalo ehamba 1 futhi ivuselele amafreyimu ehamba 2. Imiphumela ibukeka ihlala iqinile, ngisho uma uyisho uhambe endaweni engalingani.

Yebo. Ukuqinisa kunzima kakhulu ku-ffmpeg okungenani Free.ai (ama-pass amabili + uku-encode kabusha ku ~0.4× isikhathi sangempela) ngakho-ke inani liyi-80 tokens ngesekondi ye-source ne-500-token floor. I-30-second clip ibiza ama-2,400 tokens — ngaphakathi kwe-pool yansuku zonke yabasebenzisi ababhalisiwe ngosuku olujwayelekile.

vid.stab iyilayibrari efanayo yokulinganisela i-motion engezansi kwe-OpenShot, Avidemux, kanye neminye imijikelezo yomsebenzi ye-pro. Ilandelela ama-rasters angama-100 ngefreyimu ngayinye futhi ithuthukisa indlela yekhamera elula, ngenkathi izihlungi ezingenalutho zibheka kuphela izinkinga ze-pixel. I-output ibukeka ilula — ikakhulukazi kumavidiyo ahambayo.

Subtle, Medium, and Maximum. Subtle isebenzisa ifasitela elincane lokugcoba (ifreyimu engu-10) kanye nokugcoba okuncane — okulungile ku-footage elungile oyifunayo ukuwuhlanza. Medium (ifasitela le-20-frame, 2% zoom) kuyiphutha. Maximum (ifasitela le-40-frame, 4% zoom) ku-drunk-cam-level shake lapho uvumela khona ukugcoba okukhulu ukuthola isithombe esiqinile.

Ukuqinisa kufanele kukhulule kancane ukuze kufihla iziphetho eziphuma ngaphandle kwefremu eqinisiwe. Okuncane kulahlekelwa ~1%, Okuphakathi ~2%, Okuphezulu ~4%. Khetha indlela yokucropa yokukhulula-ukugcwalisa uma ufuna ukuphuma ku-resolution efanayo ne-input (iziphetho zigcwele ngokucacile ngokukhulula).

MP4, MOV, WebM, MKV, AVI — noma iyiphi ibhokisi elisha le-ffmpeg-decodable. I-output ihlala i-H.264 MP4 ukuze kube lula ukuhambisana. I-input cap yi-200MB; ama-clip ancane asheshisa futhi abiza ngaphezu.

Isikhathi esigcwele 2-3×. I-clip yemizuzu engu-30 iphele ngemizuzu engu-60-90. Ama-clips ade kakhulu angafika ku-ffmpeg timeout yethu yemizuzu engu-8 — uma une-vlog eshayayo imizuzu engu-5, cindezela i-60-second chunks kuqala.

Yebo — isithombe se-drone siyisibonelo esifanele sokusebenzisa i-vid.stab. Ukuqina okuphakathi kunciphisa ukungqubuzana kwe-gimbal kanye nokungqubuzana kwe-prop ngaphandle kokushaya ifreyimu. Ungagcini ngokuqinisa isithombe esikhona esisebenza ngokulungisa i-gimbal; uzophinde ulungisa futhi uveze ukungqubuzana.

Yebo. Umsindo udluliswa ngaphandle koshintsho (i-stream-copy) ngakho-ke akukho lutho oluzophinde lukhodwe noma ushintsho lokuhlanganiswa. Uhlelo lwevidiyo kuphela oluzophinde lushintshwe.

Ukulungisa yi-NLE ephelele ene-stabilizer enhle engezansi kwe-1GB yokufakela. Ukulungisa oku-off-one, ukufaka isiphequluli 10x okulula. Uma uhlela iphrojekthi ye-multi-track nge-color grading, sebenzisa Ukulungisa; uma ufuna nje i-clip eyodwa eqiniswe, sebenzisa lokhu.

Akunalutho. Amafayela okungenisa asuswa uma ukuqiniswa kuqediwe. Ukuphuma okuqinisiwe kugcinwa ku-CDN yethu ngehora le-24 (izinsuku ezi-7 kubasebenzisi ababhalisiwe nge-credits).

Yebo. POST ingxenye eminingi ku /v1/video/stabilize/ nge `file`, `strength` (subtle / medium / max), `crop` (auto / no-crop / zoom). Indiza-ngaphambili: GET /v1/video/stabilize-quote/?duration=SECS. Izinhlamvu ezigcwele ku /api/.

Bhala mahhala nge-10,000 tokens

Dala i-akhawunti ekhululekile

Akukho khadi le-credit elidingekayo

Ungayilinganisela kanjani lethuluzi?

Uthanda i-Free.ai? Ngisho nabahlobo bakho!