HeyGen Not Working?
Video stuck processing, API v1 deprecated errors, voice cloning failed, credits exhausted, or lip sync off? Check live status and fix it fast.
HeyGen — live status
Updated every 5 minutes. Full history at prismix.dev/service/heygen.
What's wrong? Diagnose fast
Video stuck processing
Normal processing is 1–10 minutes. If stuck beyond 30 minutes: refresh and check "My Videos" in the sidebar — the video may have completed without updating the generation page UI. Cancel and resubmit if still stuck.
API v1 deprecated errors
HeyGen deprecated the v1 API. Change your base URL from api.heygen.com/v1/ to api.heygen.com/v2/. The v2 video.generate endpoint, avatar listing, and voice listing all have updated JSON schemas — see migration guide below.
Voice cloning failed
Instant Voice requires clean audio, no background music, 30 sec–3 min duration, MP3/WAV/M4A format under 50MB. Robot-sounding clones = re-record in a quiet room. Requires Creator plan.
Credits exhausted
Each video minute costs 1+ credits depending on avatar type. Credits do not roll over. Check balance at Settings → Credits. Purchase add-on packs or upgrade plan for more monthly credits.
Lip sync off
Lip sync accuracy depends on audio quality. For custom audio uploads: use WAV at 44.1kHz, avoid compression artifacts. For TTS: switch to a different voice model. Video translation lip sync is less accurate for languages very different from English.
Video translation not accurate
HeyGen Video Translation works best for source videos with clear speech, minimal background noise, and one speaker. Multi-speaker videos, heavy accents, or fast speech reduce accuracy. Some language pairs have better quality than others (Spanish, French, German top performers).
HeyGen API v1 → v2 migration reference
| Action | v1 (deprecated) | v2 (current) |
|---|---|---|
| Create video | POST /v1/video.generate | POST /v2/video/generate |
| Check video status | GET /v1/video_status.get | GET /v2/video_status.get |
| List avatars | GET /v1/avatar.list | GET /v2/avatars |
| List voices | GET /v1/voice.list | GET /v2/voices |
| Auth header | Authorization: Bearer {key} | X-Api-Key: {key} |
HeyGen plans and credit limits
| Plan | Price | Credits/mo | Instant Voice | API access |
|---|---|---|---|---|
| Free | $0 | 1 credit (1 min video) | No | No |
| Creator | $29/mo | 15 credits/mo | Yes | No |
| Business | $89/mo | 30 credits/mo | Yes | Yes |
| Scale | $199/mo | 100 credits/mo | Yes | Yes (higher limits) |
| Enterprise | Custom | Custom | Yes | Yes (dedicated) |
1 credit = 1 minute of video on standard avatars. Premium avatars cost more credits per minute. Credits do not roll over month to month.
Step-by-step fix
- 1
Check live status
Visit prismix.dev/service/heygen. HeyGen video rendering, voice cloning, and translation are separate systems. If only voice cloning fails, check if that specific feature is listed as degraded.
- 2
Fix stuck video processing
Navigate to My Videos in the left sidebar — completed videos appear here even if the generation page didn't update. If the video is missing or shows "Failed": cancel it, then re-create the scene with the same script and avatar. If you were charged credits for a failed video, contact HeyGen support at support.heygen.com for a credit refund.
- 3
Migrate from API v1 to v2
Update your API base URL to
https://api.heygen.com/v2/. Change the auth header fromAuthorization: Bearer {key}toX-Api-Key: {key}. Update video generation payload: v2 usesvideo_inputs[].characterobjects instead of flat avatar_id fields. See the migration table above. - 4
Fix voice cloning issues
Re-record your audio sample: (1) record in a quiet room — no background music, HVAC hum, or echoes; (2) speak naturally for 2–3 minutes; (3) use a good microphone (built-in laptop mics work but an external USB mic improves quality); (4) save as WAV or MP3 and verify the file is under 50MB; (5) upload and wait 3–5 minutes for processing. The quality indicator in the Voice Lab shows if the sample is sufficient.
- 5
Fix lip sync issues
For TTS voice: try different voice models — some have better lip sync accuracy. Avoid very fast speech rates (use 90–110% speed). For uploaded audio: use clean WAV at 44.1kHz, avoid heavy compression. If the video was created correctly but lips are slightly off, use the "Adjust Lip Sync" slider in the editor to add a delay offset.
- 6
Check and replenish credits
Go to
app.heygen.com→ click your avatar (top right) → Settings → Credits. Check your balance and next renewal date. Credits do not roll over — unused credits expire on your monthly billing date. Purchase add-on credit packs from the same Credits page if needed before your renewal.
Get alerted when HeyGen goes down
Star HeyGen on Prismix and get emailed the moment status changes. Free, no credit card.
Frequently asked questions
Why is HeyGen not working?
Common HeyGen issues: video stuck in processing (refresh, check My Videos), API v1 deprecated (migrate to v2 base URL), credits exhausted (Settings → Credits), voice cloning failed (clean audio 30s–3min required), or a platform outage (check prismix.dev/service/heygen).
Is HeyGen down right now?
Check prismix.dev/service/heygen for live HeyGen status. Video generation, voice cloning, and video translation are separate systems — a partial outage may only affect one feature.
HeyGen API error — v1 deprecated, how to fix?
HeyGen deprecated v1 in favour of v2. Change base URL to api.heygen.com/v2/, change auth header from "Authorization: Bearer {key}" to "X-Api-Key: {key}", and update your request body to the v2 schema (video_inputs array with character objects). Key v2 endpoints: POST /v2/video/generate, GET /v2/video_status.get, GET /v2/avatars, GET /v2/voices.
HeyGen video stuck in processing — what to do?
Processing normally takes 1–10 minutes. Check "My Videos" in the left sidebar — completed videos appear there even if the page doesn't update. If stuck beyond 30 minutes, cancel and resubmit. Contact HeyGen support at support.heygen.com if you were charged credits for a failed video.
HeyGen voice cloning not working — how to improve quality?
Instant Voice Cloning requires: clean audio (no background noise or music), 30 seconds to 3 minutes of the target speaker, MP3/WAV/M4A format, under 50MB file size. 2–5 minute samples produce better clones. Record in a quiet room with a USB microphone for best results. Creator plan or above required.
