Compare AI video generation models and pricing
A prompt in. A finished clip out.
Explore video generation models for short clips, product creative, animation, and media workflows. Models are ranked only when their duration and billing units are comparable.
73 video models are listed today. All of them use one OneInfer API key and pay-as-you-go wallet.
Choose and call a video model in 4 steps
The list ranks them by price per generated second, with flat per-video prices in their own group. Different billing units are never mixed, so the order is fair.
Lowest comparable price: Kuaishou: Kling Lip Sync
Currently the lowest-priced option in the provider price per generated video second group. Different billing units are never mixed.
$0.0028 / second
View pricing and providersAvailable video models
Comparable models are ordered by their current normalized price. Pricing content refreshed November 9, 2026.
73 models
| Model |
|---|
 bytedance ByteDance lightweight, cost-efficient video generation model
|
bytedance ByteDance ultra-fast video generation model for high-throughput pipelines
|
bytedance ByteDance next-generation cinematic video generation model with synchronized audio
|
deepinfra FastWan 2.2 5B makes a 5-second 480p or 720p clip from a prompt, quickly and cheaply.
|
deepinfra FastWan 1.3B makes a 5-second 480p clip from a prompt in seconds, for very cheap drafts.
|
 Kuaishou: Kling Lip SyncLowest pricefal Kling Lip Sync drives the speaker in a 2-10 second 720p or 1080p video with an audio track of up to 60 seconds. Billed per started 5 seconds of video.
|
 minimax MiniMax-Hailuo-02 is a state-of-the-art AI video generation model capable of producing high-fidelity 1080p video (up to 10 seconds) from text or images. It is recognized for exceptional, realistic physics simulations (fluid dynamics, collision) and strong character consistency. It supports both Text-to-Video (T2V) and Image-to-Video (I2V) capabilities.
|
minimax MiniMax Hailuo 2.3 Fast is an image-to-video generation model optimized for low-latency and computational efficiency. Generates 6-10 second videos at 768p (up to 1080p) with 30-50% faster generation than standard model. Ideal for rapid iteration, social media content, and bulk video production.
|
minimax MiniMax-Hailuo-2.3 is a high-fidelity AI video generation model specializing in cinematic realism, superior character physics, and complex motion, available in standard and fast versions. It supports 768p or 1080p resolution, creates 6-10 second clips, and excels at text-to-video and image-to-video (I2V) tasks with high prompt adherence.
|
 deepinfra Wan2.2-T2V-A14B is a 14B-parameter text-to-video diffusion model specializing in high-fidelity motion generation and temporal consistency. Features advanced motion dynamics modeling and cinematic quality rendering with 128-frame coherence support.
|
novita Wan2.2-I2V-A14B is a 14B-parameter image-to-video diffusion model specializing in animating static images with realistic motion. Features advanced motion transfer, temporal consistency, and style preservation capabilities with 128-frame video generation from single images.
|
novita Seedance V1 Lite is a lightweight text-to-video generation model capable of producing short video clips from textual prompts.
|
novita Seedance V1.5 Pro I2V is an advanced image-to-video generation model that animates input images with superior temporal coherence, higher resolution, and improved motion quality.
|
novita Seedance V1.5 Pro T2V is an advanced text-to-video generation model featuring improved temporal consistency, higher resolution output, and enhanced prompt adherence.
|
 grok Grok Imagine Video 1.5 Lite is the low-cost version of xAI's Grok Imagine Video 1.5, at a quarter of its price per second. It makes videos from a prompt, or animates a starting image.
|
deepinfra P-Video is Pruna AI's fast video model: 1 to 10-second clips at 720p or 1080p from a prompt, a start image or an audio track.
|
deepinfra PixVerse V6 makes 1 to 15-second clips from a prompt or a start image, from 360p to 1080p, with optional sound.
|
deepinfra PixVerse V6 animates a start image for 1 to 15 seconds, from 360p to 1080p, with optional sound.
|
deepinfra P-Video Avatar makes a portrait speak: the input audio, or the prompt read in a generated voice.
|
 lightricks LTX-2.3 Fast is a speed-optimized variant of Lightricks' LTX-2.3 video generation model. It is designed for high-speed, cost-effective generation of synchronized audio and video in a single pass. It supports up to 20-second clips, native 9:16 portrait and 16:9 aspect ratios, and resolutions up to 4K. Delivering significantly faster inference than the Pro version, it is ideal for rapid prototyping, batch generation, and short-form social media content.
|
 Gemini Omni 1.1 Flash is Google's high-performance multimodal video generation and editing model designed for fast, conversational video creation and cinematic control. It supports text-to-video, image-to-video, video-to-video editing, natural-language conversational editing, video extension, first-and-last-frame interpolation, and resolution upscaling. The model combines multimodal understanding with generative video capabilities to produce coherent, controllable video with synchronized audio. It supports a 1,048,576-token context window and generates videos lasting 3 to 10 seconds at 24 FPS, with output resolutions ranging from 360p to 4K. It is suitable for filmmaking, advertising, animation, social media content, creative production, and iterative video editing workflows.
|
lightricks LTX-2.3 is a DiT-based audio-video foundation model designed to generate synchronized video and audio within a single model. It brings together the core building blocks of modern video generation, with open weights and a focus on practical, local execution.
|
fal Wan 3.0 Video is Alibaba's high-speed AI video generation model from the Wan 3.0 family. It delivers the core capabilities of the standard Wan 3.0 Video model with significantly faster end-to-end generation speeds, making it ideal for high-volume API workflows and production. It can generate up to 30 seconds of synchronized audio and video in a single pass (at 480p, 720p, or 1080p resolution). The model supports extensive multimodal references, natively accepting text, images, audio, video.
|
 fal Vidu Q4 (preview) animates a start frame into a 3-16 second video at up to 4K, with dialogue and sound effects generated alongside the picture. Billed per second of requested video.
|
fal Vidu Q4 (preview) keeps the people and objects of up to 12 reference images, and the voices of up to 3 audio clips, consistent across a 3-16 second video at up to 4K, with optional sound. Refer to them in the prompt as [@reference_image_1] and [reference_audio_1]. Billed per second of requested video.
|
grok Grok Imagine Video is xAI's foundational multimodal video generation model capable of turning text prompts and static images into dynamic video clips. It supports high-quality visual motion, cinematic camera controls, and clip lengths of up to 10 seconds, establishing the baseline for xAI's video generation capabilities.
|
 oneinfer Veo 3.1 Lite is Google's cost-efficient video generation model designed for high-volume applications and rapid iteration. It generates video from text or image prompts with synchronized audio and supports cinematic controls, multiple aspect ratios, and short-form production workflows.
|
minimax MiniMax H3 Max is MiniMax's video generation model designed for text-to-video and image-to-video generation. It supports generating videos directly from text prompts as well as using first-frame and/or last-frame images to control the beginning and ending of generated sequences. The model produces videos at 480P or 768P resolution with configurable durations from 5 to 15 seconds and supports common or adaptive aspect ratios.
|
deepinfra Removes a video's background, keeping the subject. Trained only on licensed data.
|
deepinfra Upscales a video to two times its resolution.
|
deepinfra Erases what a mask video marks (the second input video) from a video and fills in the background.
|
fal Kling AI Avatar v2 Standard turns one image of a person, animal, cartoon or stylized character and an audio track into a talking video, with an optional prompt. Billed per second of output video.
|
fal MiniMax H3 is a general-purpose, omni-modal generative system that supports unified understanding of multimodal contexts composed of text, images, video, and audio. It can generate video with native stereo audio at resolutions up to 2K and durations of up to 15 seconds. It is released with open weights under the MiniMax H3 Community License.
|
lightricks LTX-2.3 Pro Audio-to-Video makes a video timed to an audio track: dialogue, music or ambient sound. The audio sets the length (2 to 20 seconds at 720p and 1080p, up to 10 at 1440p and 4K) and is the video's sound track. Describe the video in the prompt, give it a start frame (and an end frame), or both. Send the audio in audio_urls. Billed per second of audio.
|
bytedance ByteDance high-fidelity video generation model supporting multi-aspect ratio output
|
fal Wan 3.0 Prime is Alibaba's accelerated multimodal video generation model designed for high-quality video creation with significantly faster generation speed. It supports text-to-video, image-to-video with first-frame or first-and-last-frame conditioning, and reference-to-video generation using multimodal references including text, images, video, and audio. The model supports native audio generation, enhanced reasoning for complex prompts, multiple aspect ratios, resolutions up to 1080p, and video durations of up to 30 seconds.
|
fal Aurora from Creatify turns one image and an audio track into a speaking or singing avatar video at 480p or 720p, with an optional prompt. Billed per second of output video, rounded up.
|
grok Grok Imagine Video 1.5 is xAI's advanced video model built on the Aurora-2 engine. Its standout feature is native one-pass audio generation, which produces synchronized dialogue (lip-sync), sound effects, and background music simultaneously with the video. Supporting resolutions up to 1080p and durations up to 15 seconds, it allows for complex workflows including text-to-video, image-to-video, video extension, and multi-image reference guidance to maintain consistent styles and characters.
|
fal Fabric 1.0 from VEED animates one image into a talking video from an audio track, at 480p or 720p. Billed per second of output video.
|
fal Lipsync 2 Pro from Sync Labs matches the speaker's lips in an existing video to a new audio track, with detail-preserving face rendering. Billed per second of output video.
|
novita Kling v3.0 Standard T2V is Kuaishou's text-to-video generation model designed to create cinematic video directly from natural-language prompts. It supports smooth motion, strong prompt adherence, multi-shot composition, negative prompting, configurable aspect ratios, flexible video durations, and optional native audio generation for synchronized audiovisual content.
|
novita Kling v3.0 Standard I2V is Kuaishou's image-to-video generation model designed to animate still images with realistic motion and cinematic scene dynamics. It supports prompt-guided animation, first-frame and optional end-frame conditioning, configurable generation strength, negative prompting, multi-shot composition, flexible video durations, and optional native audio generation.
|
lightricks LTX-2.5 Fast is an accelerated endpoint for Lightricks' LTX-2.5 video generation model. Designed for high-speed text-to-video and image-to-video workflows, it trades a degree of rendering fidelity for significantly faster inference speeds. It supports up to 20-second clips, 4K resolution output, and native synchronized audio generation, making it ideal for rapid prototyping and high-throughput production.
|
lightricks LTX-2.5 Fast Audio-to-Video makes a video timed to an audio track: dialogue, music or ambient sound. The audio sets the length (2 to 20 seconds at 720p and 1080p, up to 10 at 1440p and 4K) and is the video's sound track. Describe the video in the prompt, give it a start frame (and an end frame), or both. Send the audio in audio_urls. Billed per second of audio.
|
oneinfer Veo 3.1 Fast is Google's speed-optimized video generation model balancing generation quality, latency, and cost. It creates high-quality video from text or image prompts with native synchronized audio and supports first-frame and last-frame conditioning.
|
deepinfra Wan 2.6 T2V is Alibaba's text-to-video generation model designed to create high-quality cinematic video directly from natural-language prompts. It supports complex scene descriptions, realistic motion and physical dynamics, single-shot and multi-shot generation, prompt enhancement, negative prompting, multiple aspect ratios, high-resolution video generation, and optional native audio for synchronized audiovisual output.
|
novita Wan 2.6 V2V is Alibaba's video-to-video generation model designed to transform reference videos while preserving important character, motion, and structural information. It supports prompt-guided character replacement, role-playing, visual style and texture transformation, multiple reference videos, single-shot and multi-shot generation, prompt enhancement, negative prompting, high-resolution output, and synchronized audio-video generation.
|
lightricks LTX-2.3 Pro Retake regenerates one section of an existing video (its picture, its sound or both) and keeps the rest. Send the video in video_url, the section's start in extra_body.start_time and its length in duration, and describe the new section in the prompt; extra_body.mode is replace_audio_and_video (default), replace_video or replace_audio. The video must be at least 73 frames (about 3 seconds). Output is 1080p. Billed per second of the input video.
|
lightricks LTX-2.3 Pro Extend continues a video by 2 to 20 seconds, matching its motion and sound. Send the video in video_url and the seconds to add in duration, and describe them in the prompt; extra_body.mode is end (default) or start, and extra_body.context sets how many seconds of the video it reads. Output is 1080p. Billed per second added plus the seconds it reads, about 21 seconds at most.
|
lightricks LTX Reframe changes a video's aspect ratio, generating the parts of the frame the video doesn't fill and keeping the whole video. Send the video in video_url (up to 60 seconds) and pick the output with resolution (720p or 1080p) and aspect_ratio (1:1, 4:5, 5:4, 9:16 or 16:9); extra_body.placement positions the video in the frame. Billed per second of the input video, by the output resolution.
|
fal Fabric 1.0 Fast from VEED animates one image into a talking video from an audio track, faster than Fabric 1.0, at 480p or 720p. Billed per second of output video.
|
Gemini Omni Flash is Google's multimodal video generation and editing model designed for fast, conversational video creation. It supports text-to-video, image-to-video, video-to-video editing, video extension, keyframe interpolation, resolution upscaling, and natural-language iterative editing through the Gemini Interactions API. The model generates videos with synchronized audio and supports a 1,048,576-token context window. It can generate videos lasting 3 to 10 seconds at 24 FPS, with supported output resolutions ranging from 360p to 4K. It is suitable for advertising, filmmaking, animation, social media production, and creative video workflows.
|
fal OmniHuman 2.0 from ByteDance turns one image and an audio track of 4 seconds to 5 minutes into a 720p talking, gesturing video, for presenters, explainers, marketing and music videos. An optional prompt guides the scene, movement and camera. Billed per second of output video.
|
deepinfra Wan2.7 animates a start frame (and optional end frame), or continues a clip, for 2 to 15 seconds at 720p or 1080p, with optional audio driving the motion.
|
deepinfra Wan2.7 R2V makes a video featuring the characters or scenes in up to five reference images or a reference video, at 720p or 1080p.
|
novita Kling v3.0 Pro T2V is Kuaishou's professional-grade text-to-video model optimized for high-fidelity cinematic generation from natural-language prompts. It provides enhanced motion quality, prompt adherence, subject consistency, multi-shot storytelling, complex scene generation, negative prompting, flexible video durations, and optional native audio generation for professional audiovisual workflows.
|
novita Kling v3.0 Pro I2V is Kuaishou's professional-grade image-to-video model for transforming still images into high-fidelity cinematic video. It offers enhanced motion generation and subject consistency with prompt-guided animation, first-frame and optional end-frame control, multi-shot composition, negative prompting, configurable generation strength, flexible durations, and optional native audio generation.
|
fal Kling AI Avatar v2 Pro turns one image of a person, animal, cartoon or stylized character and an audio track into a higher-quality talking video, with an optional prompt. Billed per second of output video.
|
lightricks LTX-2.5 Pro is the quality-optimized API endpoint of Lightricks' LTX-2.5 audio-video model. It utilizes Diffusion Fidelity Rendering to allocate maximum compute to complex scenes, ensuring high pixel quality, sharp faces, legible text, and seamless multi-shot continuity. It is designed for professional text-to-video and image-to-video workflows with synchronized native audio.
|
lightricks LTX-2.5 Pro Audio-to-Video makes a video timed to an audio track: dialogue, music or ambient sound. The audio sets the length (2 to 10 seconds) and is the video's sound track. Describe the video in the prompt, give it a start frame (and an end frame), or both. Send the audio in audio_urls. Billed per second of audio.
|
fal Lipsync 3 is Sync Labs' current lip-sync model. It re-animates the speaker's mouth in an existing video to match a new audio track, keeping the rest of the frame, at up to 4K. Billed per second of output video.
|
deepinfra Makes a mask video of a video's foreground subject.
|
deepinfra Makes a mask video of what a prompt describes in a video.
|
fal Kling 4.0 turns a prompt into a 3-30 second video at 720p or 1080p, with a stereo sound track generated alongside the picture. Billed per second of requested video.
|
fal Kling 4.0 animates a start frame, optionally towards an end frame, into a 3-30 second video at 720p or 1080p with a generated stereo sound track. Billed per second of requested video.
|
fal OmniHuman 1.5 from ByteDance turns one portrait image and an audio track into a talking, gesturing video, with an optional prompt to guide the performance. Audio up to 30 s at 1080p or 60 s at 720p. Billed per second of output video.
|
 blackforestlabs FLUX 3 Video is Black Forest Labs' multimodal video generation model capable of producing high-quality video with synchronized speech, sound effects, and environmental audio. It supports text-to-video generation, image-to-video animation, multi-keyframe conditioning, video-to-video continuation, and multi-scene cinematic sequences. The model generates clips lasting 5 to 20 seconds at 24 FPS, with HD or Full HD output, and supports multilingual speech with lip synchronization, realistic physical motion, expressive characters, animation, and stylized visual content. It also provides draft generation and enhancement workflows for efficient creative iteration.
|
oneinfer Veo 3.1 is Google's state-of-the-art cinematic video generation model built for maximum visual fidelity. It generates high-quality video from text or image prompts with native synchronized dialogue, ambient effects, and sound, while supporting advanced cinematic controls and production workflows.
|
lightricks LTX HDR converts an SDR video to HDR and returns a ZIP of per-frame EXR images at the video's own resolution, for color grading and HDR finishing. Send the video in video_url: up to about 7 seconds at 1080p, 4 at 1440p or 2 at 4K (181, 101 or 41 frames). Billed per second of the input video, by its resolution.
|
fal Kling v3.0 4K is Kuaishou's premium text-to-video model, generating native 4K cinematic videos from text prompts. It supports 3-15 second durations, flexible aspect ratios, optional synchronized audio, and multi-prompt scene transitions for complex compositions.
|
fal Kling v3.0 4K is Kuaishou's premium image-to-video model, generating up to 15 seconds of native 4K video (30fps) from images. It supports optional synchronized audio, multi-prompt scene composition for complex narratives, and flexible aspect ratios (16:9, 9:16, 1:1).
|
deepinfra PixVerse text-to-video makes a 5 or 8-second clip from a prompt, at a flat price per video.
|
deepinfra PixVerse text-to-video in HD: a 5 or 8-second clip from a prompt, at a flat price per video.
|
How to choose
- Short-form video
- Marketing creative
- Animation
- Product visualization
Pricing methodology
OneInfer compares only positive prices with the same billing unit. Per-minute, per-character, per-token, per-image, per-video, and per-second rates remain separate. Prices can change, so the current model page and console remain the source of truth.
Frequently asked questions
How are video prices compared?
Per-second pricing is compared with per-second pricing. Flat per-video prices form a separate group because duration and resolution can change the effective cost.
Can a video model generate audio?
Some can. Native audio support is displayed only when the pricing or model metadata confirms it.