curl --request POST "https://gptproto.com/api/v3/kling/kling-v3-omni-pro/text-to-video" \
--header "Authorization: Bearer $GPTPROTO_API_KEY" \
--header "Content-Type: application/json" \
--data '{
"prompt": "A tiny origami fox sailing a teacup across a moonlit puddle",
"sound": false,
"aspect_ratio": "16:9",
"duration": 5
}'Parti dal costo di un singolo campione e scegli un budget di test. Le tariffe GPTProto sono inferiori del 20% rispetto al prezzo di listino.
| Scenario | Listino ufficiale Kling | OpenRouter | GPTProto | Risparmi al mese |
|---|---|---|---|---|
| Brevi annunci di prodotto30 × 5s/mese | $16.80 | $17.72 | $13.44 | −$3.36≈ $40.32 / anno |
| Clip per i social60 × 5s/mese | $33.60 | $35.45 | $26.88 | −$6.72≈ $80.64 / anno |
| Generazione in blocco100 × 5s/mese | $56.00 | $59.08 | $44.80 | −$11.20≈ $134.40 / anno |
Kling · $0.2688 / al secondo
Access Kling v3 Omni Pro API on GPTProto for product ads, recurring characters and reference-driven video generation. Use one API key and a shared balance across Kling and other AI models, with task-specific controls for generating new scenes or revising existing footage.
Control the Opening and Ending Frames
Animate an approved product photo as the opening frame, with an optional final image to guide the ending. Describe the action and camera movement while keeping your starting composition grounded in the original asset.
Guide Scenes with Multiple References
Supply reference images for characters, products or locations instead of rebuilding every detail in a prompt. Use video references when an existing clip should guide the new scene’s movement or visual direction.
Generate Dialogue and Scene Sound
Add dialogue, ambience and sound effects during supported video generation. For video-input tasks, retaining the original soundtrack is a separate option from creating new native audio; choose the setting that matches your source material.
Revise Existing Video with Instructions
Use existing footage as the editing base and describe the changes you want, such as a different background, object or visual style. Reference images can guide the revision without rebuilding the whole scene from scratch.
Kling v3 Omni Pro is GPTProto’s professional-tier access to Kling VIDEO 3.0 Omni, a video model from Kuaishou. Its reference workflow lets you supply visual assets for the subject, setting and style, while the prompt describes what should happen. This is useful when an ad must feature a particular product or a story must reuse the same character.
Choose the input mode according to what is already decided: the opening composition, the subject’s appearance, the movement to reference, or the footage to edit. These controls guide the result; inspect product proportions, labels and character details before using a clip.
| Specification | Details |
|---|---|
| Developer | Kuaishou / Kling AI |
| GPTProto model identifier | kling-v3-omni-pro |
| Official model identifier and tier | kling-v3-omni, professional mode |
| Pro output resolution | 1080p; GPTProto lists Omni 4K separately |
| New clip duration | 3–15 seconds; editing duration depends on the source workflow |
| Input workflows | Text-to-video, image-to-video, reference-to-video and video editing |
| Aspect ratios for text/reference generation | 16:9, 9:16 and 1:1 |
| Audio | Optional native audio in supported generation modes; source-audio retention for video-input workflows |
| Starting material | Choose | Practical use |
|---|---|---|
| A written scene brief | Text-to-video on this page | Explore an ad concept before selecting visual assets |
| An approved opening image | Image-to-video | Animate a catalog photo or guide a transition with an ending image |
| Several subject images or a guiding clip | Reference-to-video | Reuse a product or character in a new setting |
| Footage that needs changes | Video-to-video | Revise the background, objects or style of an existing shot |
An ending image defines where a clip should finish. A reference image describes what a subject or scene should look like. A reference video guides a new generation; an editing base is footage you want to change. Match the task to that intention before preparing your assets.
For a product listing, start from an approved photo when the opening composition matters. For campaign variations, use references and change one element at a time: the setting, camera movement or action. Keep the product’s shape, materials and colors explicit in the brief.
A useful first test is a short clip with one camera move and one subject action. Compare the result against the original asset, especially small labels and moving edges. Add exact prices, promotional copy and final captions in your editor so they remain consistent across campaign versions.
Example brief: Animate the supplied serum bottle on a pale stone surface. Make a slow camera push toward the bottle as soft side light moves across the glass. Keep the bottle shape, cap, label placement and liquid color unchanged. No hands, extra products or added text.
Choose by the assets and controls your task needs. The following models are available on GPTProto; these are workflow recommendations, not benchmark rankings.
| Model | Relevant input workflow | When to consider it |
|---|---|---|
| Kling v3 Omni Pro | Text, first/last frames, image/video references and editing | You need several ways to guide a product or character scene |
| Seedance 2.0 | Text, image, video and audio references, depending on the task | Your brief combines visual references with separate audio assets |
| Seedance 2.0 Mini | Text-to-video on the linked page; image-to-video has a separate task page | You are evaluating 480p/720p drafts before choosing a final direction |
| Hailuo 2.3 Pro | Image-to-video on the linked page | You want to animate one image at 1080p and add audio separately |
For a fair test, use the same brief, duration and target format. Count how many outputs meet your acceptance criteria, then compare the cost per usable clip. A lower advertised rate does not establish better value if the task requires more revisions.
Guide, confronti e aggiornamenti relativi a questo modello.
Tutti gli articoli
See which models fit dialogue, action and character continuity in short dramas. Compare Seedance, Kling, MiniMax, Vidu, Veo and Hailuo.
Learn how to make an AI cat dance video from your own photo using Wan 3.0, a reference dance clip, copy-ready prompts, and simple fixes.
Compare 5 affordable AI video APIs for ecommerce and AI short drama. See current pricing, clip costs, audio fees, and the best model for each job.
Input
Output
