Kwaivgi Kling V2.6 Pro Text To Video

Kwaivgi Kling V2.6 Pro Text To Video

Playground

Kling 2.6 Pro delivers top-tier text-to-video generation with smooth motion, cinematic visuals, strong prompt adherence, and native audio for ready-to-share clips. Ready-to-use REST inference API, best performance, no cold starts, affordable pricing.

Features

Kling 2.6 Audio Text-to-Video turns a text prompt directly into a fully scored clip: camera motion, character action, and soundtrack (voice, ambience, SFX) are generated in one pass, so the scene looks and sounds like it belongs together.

🌟 Model Highlights

• Joint audio–video generation – Visuals and sound are created together, not bolted on after the fact.
• Character-aware voices – Speech that matches who’s on screen, with timing aligned to the action you describe.
• Scene-driven sound design – Ambient noise and effects that follow the camera and events in the shot.
• Script-to-scene pipeline – Start from a natural-language prompt; Kling handles shots, motion, and soundscape.

🧩 Parameters

• prompt* – Describe what happens in the scene: characters, camera moves, environment, and audio mood (e.g. “Close-up of a robot repairing a neon sign, soft synthwave music, quiet city ambience, no dialogue.”)

• negative_prompt – Things to avoid in both visuals and audio (logo, watermark, heavy text, glitch, noise).

• cfg_scale – Guidance strength (default 0.5):

• Lower → looser, more organic; model improvises more.

• Higher → closer to prompt wording; can look or sound more “forced”.

• sound –

• On → generate video with audio (voice / ambience / SFX where appropriate).

• Off → silent video only (cheaper, same visuals).

• duration – 5 s or 10 s clips.

🎯 Typical Use Cases

• Social ads or launch teasers with built-in narration and sound design.
• Short story beats, animatics, or previz where visual + audio timing must line up.
• Product explainers with spoken description + on-screen action.
• Cinematic posts and shorts where you want music, ambience, and motion from a single prompt.

💰 Pricing

🚀 How to Use

1. Write a prompt describing:

• what the camera sees (shots, motion, setting),
• what characters do,
• and, if sound is on, the voice tone, music style, and ambience/SFX you want.

2. (Optional) Add a negative_prompt for things you don’t want in either image or audio.
3. Tune cfg_scale (start from 0.5; increase only if it’s not following your prompt enough).
4. Toggle sound on/off depending on whether you need audio.
5. Run the model.

🔎 Tips

• Write prompts like a mini shot list + audio brief: who, where, camera, mood, and sound.
• For clearer narration, explicitly specify “single narrator”, voice gender/age, and language/accents.
• Use negative_prompt for “watermark, text, logo, glitch, noisy audio” to keep outputs clean.
• For platform export (Reels/Shorts/TikTok), pick 9:16; for YouTube/web, use 16:9; for feeds/ads, try 1:1.

Authentication

For authentication details, please refer to the Authentication Guide.

API Endpoints

Submit Task & Query Result

set -euo pipefail

export WAVESPEED_API_KEY="your-api-key"

REQUEST_BODY=$(cat <<'JSON'
{
  "prompt": "A cinematic ocean wave at sunrise, highly detailed",
  "cfg_scale": 0.5,
  "sound": true,
  "aspect_ratio": "1:1",
  "duration": 5
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/kwaivgi/kling-v2.6-pro/text-to-video" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}")

TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL=$(printf '%s' "${TASK}" | jq -r '.urls.get // empty')
if [ -z "${RESULT_URL}" ]; then RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"; fi

# 2. Poll until the prediction finishes.
while true; do
  RESPONSE=$(curl --silent --show-error --fail-with-body \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}")
  RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    created|processing) sleep 2 ;;
    *) printf 'Unexpected status: %s
' "${STATUS}" >&2; exit 1 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Request Parameters

Result Response Parameters