CRITICALFailure Reference

Pika Labs Lip Sync Failure — Refund Guide

Technical Classification

Phoneme-Viseme Mapping Failure

Phoneme-Viseme Mapping Failure occurs when Pika's lip-sync module produces mouth geometry uncorrelated with the input audio. The model has learned generic mouth-movement priors but fails to align them with the specific phoneme sequence in the audio track. Closed mouths appear on vowels that require open shapes; mouth stays still through entire words; tongue position is wrong for fricatives and dentals.

How to identify this failure

✕Mouth closed during open-vowel phonemes (a, e, o)
✕Mouth stays still through entire spoken words
✕Lip rounding wrong for "oo" or "oh" sounds
✕Teeth position incorrect for fricatives (f, v, s, z)
✕Mouth movement timing drifts from audio over the clip

Real generation examples

Prompt used

"Narrator speaking the word \"hello\" directly to camera, audio attached"

Failure observed @ 0:00 → 0:01

Mouth stays closed through "he-llo" duration; brief opening at 0:01 unrelated to phoneme

Prompt used

"Person reading a paragraph from a book, lip-sync to attached audio"

Failure observed @ full duration

Mouth movement is generic chewing motion, no correlation to phoneme sequence

Documentation strength

If you need to escalate

HIGH — Pika markets lip-sync as a paid-tier feature. Documented phoneme mismatch is grounds for refund.

AVA is a pre-purchase prevention tool, not a post-purchase recovery tool. Platforms generally do not guarantee credit refunds for output-quality failures; goodwill credits are at each platform's discretion. The strength rating reflects how well-formed your support ticket can be, not a promised outcome.

Prevention + documentation steps

01
Score your prompt before you generate
Run your prompt through AVA's pre-flight scoring against the Phoneme-Viseme Mapping Failure pattern. Green light = generate. Yellow/red = rewrite using the suggested fix before you commit credits.
02
Capture Generation ID + timestamp if it failed anyway
Find the Generation ID in the URL or share link. Note the exact time when the Phoneme-Viseme Mapping Failure first appears (e.g. "failure first visible at 1.2s"). Timestamped evidence is significantly stronger than a general complaint.
03
Use the correct technical term in your support ticket
Describe this failure as "Phoneme-Viseme Mapping Failure". This term maps to a recognised internal workflow in the support system and routes the ticket to the right team.
04
Submit via the correct support channel
Runway has no direct email intake. Pro+ plan: open the in-app AI Assistant (help widget bottom-right of app.runwayml.com), describe the failure with the technical term, attach evidence. Free/Standard plan: human support isn't available — your channel is Discord #community-help with @On Call - Moderators.

Frequently asked questions

Does Pika support response lip-sync failures?

Yes — lip-sync is a paid-tier marketed feature on Pika. Support refunds when the user can demonstrate audio-phoneme to visual-viseme mismatch. Reference "Phoneme-Viseme Mapping Failure" in the ticket.

Why does Pika lip-sync fail?

The viseme-prediction model is trained on generic mouth shapes and fine-tuned to align with phoneme sequences. Under audio with strong accents, fast speech, or non-English phonemes, the alignment degrades.

Which Pika lip-sync prompts are highest risk?

Fast speech, non-English audio, accented speech, multi-speaker audio tracks, and dialogue with heavy fricatives or dentals. AVA flags lip-sync prompts with high phoneme complexity.

Score your prompt

Score your prompt against this failure mode in 30 seconds

Paste your prompt and the platform you intend to use. AVA returns a red/yellow/green score against this specific failure mode plus a concrete rewrite if the risk is high.

Install free All 105 failure modes

AVA Pro · founders' round

$50 for 6 months of unlimited scoring across all failure modes + personal failure-history dashboard. Locks in $13/mo grandfathered after.

Claim $50 founders