AI video models: every model, price and API

Name: AI video model directory
Creator: HoneyVid

HoneyVid compares 231 AI video models from 54 makers and 1,263 API and app offers, each linked to its source. Of the 20 top arena scorers, HeyGen Video 1 is the cheapest: $0.05 per 5-second clip, $0.01/s on each::labs (Eachlabs). Cheapest overall: FastWan 2.2 5B at $0.00225/s on DeepInfra. 145 live models animate a still image and 74 generate native audio. We checked every price on 5 November 2026.

231models
54makers
38APIs
30apps
5 Novlast checked
$0.05 / 5 s clipCheapest top-20: HeyGen Video 1

Filters

Can do

Pick several to find models that do them all. The short tags match the badges in the table.

Where to use it

Max API priceAny

Min resolution

Min clip length

Available on

Maker

Country

Status

Open weights only

Search models

221 of 231 models

Price

AI video models with capabilities, limits, cheapest API price and number of providers. Open a model to see every API and app offer.
Compare
	T2V I2V F+L Ref V2V Audio Extend Edit	1080p	30 s	$0.035/sReplicate	13	6	Aug 2026	1172
Alibaba's newest Wan: 2-30 s audio-video generation, reference, editing and extension; #1 on the Artificial Analysis text-to-video arena. API: Replicate $0.035/s, Kie.ai $0.04/s, AI/ML API $0.05/s, Alibaba Cloud Model Studio $0.05/s, DeepInfra $0.05/s, each::labs (Eachlabs) $0.05/s, fal.ai $0.05/s, Higgsfield API $0.05/s, ModelsLab $0.05/s, OpenRouter $0.05/s, Runway API $0.05/s, WaveSpeedAI $0.05/s, Segmind $0.053/s. Apps: Poe $0.07/s, Higgsfield $0.075/s (Starter $15/mo), Pollo AI $0.113/s (Lite $15/mo), Krea $0.15/s (Basic $9/mo), Figma Weave (Weavy) Starter $24/mo, OpenArt Starter $14/mo.
	T2V I2V Audio	—	—	—	0	0	Sep 2026	1169
Utopai Studios' film-focused model, 2nd on the Artificial Analysis text-to-video and image-to-video arenas; sold inside Utopai's PAI platform, no public API.
	T2V I2V F+L Ref V2V Audio Edit	4K	15 s	$0.024/sSegmind	15	7	Jul 2026	1160
Open-weight multimodal video model that takes text, image, video and audio references; top-3 on the Artificial Analysis text-to-video arena. API: Segmind $0.024/s, PiAPI $0.03/s, Kie.ai $0.04/s, fal.ai $0.05/s, MiniMax API $0.08/s, Novita AI $0.08/s, Replicate $0.08/s, each::labs (Eachlabs) $0.09/s, ModelsLab $0.096/s, Runway API $0.10/s, WaveSpeedAI $0.10/s, AI/ML API $0.13/s, Higgsfield API $0.13/s, OpenRouter $0.13/s, Together AI $0.139/s. Apps: OpenArt $0.14/s (Starter $14/mo), Poe $0.18/s, Krea $0.26/s (Basic $9/mo), Pollo AI $0.30/s (Lite $15/mo), Figma Weave (Weavy) Starter $24/mo, Hailuo AI app Entry plan $9.99/mo, Leonardo.Ai Essential $12/mo.
	T2V I2V F+L Audio	—	20 s	$0.043/seach::labs (Eachlabs)	3	1	Oct 2026	1156
Faster post-trained FLUX 3 built with fal Research and served on fal; 4th on the Artificial Analysis text-to-video arena. Not on BFL's own price list. API: each::labs (Eachlabs) $0.043/s, Segmind $0.043/s, fal.ai $0.17/s. Apps: Krea $0.343/s (Basic $9/mo).
	T2V I2V F+L Ref V2V Motion Lips Audio Extend Edit	4K	30 s	$0.103/sOpenRouter	15	13	Aug 2026	1155
ByteDance's top video model: 4-30 s audio-video clips with reference, editing and extension; #1 on the Artificial Analysis image-to-video arena. API: OpenRouter $0.103/s, Replicate $0.103/s, BytePlus ModelArk $0.103/s, Segmind $0.109/s, each::labs (Eachlabs) $0.119/s, ModelsLab $0.124/s, Kie.ai $0.14/s, PiAPI $0.15/s, WaveSpeedAI $0.18/s, Runway API $0.20/s, Higgsfield API $0.206/s, fal.ai $0.221/s, AI/ML API, Magnific API (formerly Freepik API), Together AI $0.12 per undefined s clip. Apps: CapCut $0.12/s (Pro), Poe $0.206/s, Pollo AI $0.225/s (Lite $15/mo), Invideo $0.336/s (Pro $30/mo), OpenArt $0.455/s (Starter $14/mo), Krea $0.47/s (Basic $9/mo), Dreamina (CapCut), Figma Weave (Weavy) Starter $24/mo, Freepik (Magnific) €0.33/s (Premium €15/mo), Higgsfield Plus $49/mo, Kaiber Creator $29/mo, Monica Pro $8.30/mo, Runway (app) Standard $15/mo.
	T2V I2V Ref Audio	2K	30 s	$0.01/seach::labs (Eachlabs)	6	1	Oct 2026	1153
HeyGen's general video model: picture, dialogue and sound in one pass, 480p to 2K; 6th on the Artificial Analysis text-to-video arena and 1st on its silent board. API: each::labs (Eachlabs) $0.01/s, fal.ai $0.02/s, HeyGen API $0.02/s, OpenRouter $0.02/s, Segmind $0.021/s, WaveSpeedAI $0.025/s. Apps: HeyGen app.
	T2V I2V F+L Ref V2V Lips Audio Extend Edit Camera	2K	15 s	$0.035/seach::labs (Eachlabs)	8	3	Aug 2026	1148
Extended H3 for multi-asset creation with image, video and audio references; #2 on the silent text-to-video arena. API: each::labs (Eachlabs) $0.035/s, AI/ML API $0.05/s, fal.ai $0.05/s, MiniMax API $0.05/s, Novita AI $0.05/s, OpenRouter $0.05/s, Runway API $0.05/s, Segmind $0.053/s. Apps: Krea $0.07/s (Basic $9/mo), OpenArt Starter $14/mo, Pollo AI Lite $15/mo.
	T2V I2V F+L Ref V2V Audio Extend Edit	4K	20 s	$0.15/sBFL API	9	3	Jul 2026	1139
Black Forest Labs' video model with synchronized audio: up to 20 seconds and 4K in one request, keyframes, references, continuation and a cheap draft mode. API: BFL API $0.15/s, AI/ML API $0.17/s, each::labs (Eachlabs) $0.17/s, fal.ai $0.17/s, OpenRouter $0.17/s, WaveSpeedAI $0.17/s, ModelsLab $0.204/s, Segmind $0.243/s, Together AI $0.17 per undefined s clip. Apps: Krea $0.343/s (Basic $9/mo), Pollo AI $0.36/s (Lite $15/mo), OpenArt Starter $14/mo.
	T2V I2V F+L Ref V2V Audio Extend Edit	4K	15 s	$0.067/sOpenRouter	16	15	Feb 2026	1136
Multimodal audio-video model with reference, editing and extension up to native 4K; top-3 on the Artificial Analysis image-to-video arena. API: OpenRouter $0.067/s, BytePlus ModelArk $0.07/s, AI/ML API $0.071/s, Segmind $0.074/s, Replicate $0.08/s, Kie.ai $0.095/s, ModelsLab $0.10/s, PiAPI $0.10/s, WaveSpeedAI $0.12/s, Runway API $0.14/s, Higgsfield API $0.141/s, fal.ai $0.173/s, DeepInfra, each::labs (Eachlabs), Magnific API (formerly Freepik API), Together AI $0.16 per undefined s clip. Apps: Poe $0.129/s, Krea $0.163/s (Basic $9/mo), OpenArt $0.186/s (Starter $14/mo), Adobe Firefly $0.20/s (Firefly Standard $9.99/mo), Higgsfield $0.216/s (Plus $49/mo), Pollo AI $0.30/s (Lite $15/mo), Runway (app) $0.96/s (Standard $15/mo), CapCut Pro, Dreamina (CapCut), Flora Starter $18/mo, Freepik (Magnific) €0.21/s (Premium €15/mo), Hedra Studio, Invideo Pro $30/mo, Kaiber Creator $29/mo, LTX Studio Standard $35/mo.
	T2V I2V F+L Ref V2V Audio Extend Edit Upscale	4K	10 s	$0.03/sfal.ai	10	7	Aug 2026	1129
Google's video generation and editing model: native audio, conversational edits, extensions to 40s, 360p drafts and 1080p/4K upscaled output; Google's top model on the Artificial Analysis video arena. API: fal.ai $0.03/s, WaveSpeedAI $0.03/s, each::labs (Eachlabs) $0.034/s, Gemini API (Google AI Studio) $0.034/s, Gemini Enterprise Agent Platform (Vertex AI) $0.034/s, Runway API $0.034/s, Segmind $0.042/s, Replicate $0.05/s, Kie.ai $0.079/s, AI/ML API $0.101/s. Apps: Google Flow $0.037/s (Google AI Plus $4.99/mo), Krea $0.261/s (Basic $9/mo), Freepik (Magnific) €0.18/s (Premium €15/mo), Gemini app Google AI Plus $4.99/mo, OpenArt Starter $14/mo, Poe, Pollo AI Lite $15/mo.
	T2V I2V F+L Ref Audio Upscale	4K	15 s	$0.01/seach::labs (Eachlabs)	5	0	Sep 2026	1127
Pruna's fast video model built on MiniMax H3: 5 to 15 second clips with audio, 480p/768p native and upscaled to 4K. API: each::labs (Eachlabs) $0.01/s, Segmind $0.01/s, Pruna API $0.02/s, Replicate $0.02/s, WaveSpeedAI $0.02/s.
	T2V I2V Ref Audio	—	—	—	0	0	Sep 2026	1121
Open-source accelerated MiniMax H3 (8 steps instead of 49) under the MiniMax H3 Community License; top-6 on the image-to-video arena.
	T2V I2V Ref	1080p	—	$0.12/sVideo Rebirth API	1	1	Oct 2026	1110
Video Rebirth's production model with up to 12 image, video or audio references; #2 on the silent image-to-video arena per the maker. API: Video Rebirth API $0.12/s. Apps: BACH app.
	T2V I2V Audio	—	—	—	0	0	Aug 2026	1067
Low-cost audio-video model from Agnes AI that ranks #12 on the text-to-video arena.
	T2V I2V F+L Ref Lips Audio	1080p	15 s	$0.012/sKie.ai	11	4	May 2026	1058
xAI's top video model: native 1080p, 1 to 15 second clips with lip-synced speech, keyframes, first/last frame and image plus voice references. API: Kie.ai $0.012/s, AI/ML API $0.08/s, fal.ai $0.08/s, OpenRouter $0.08/s, Replicate $0.08/s, WaveSpeedAI $0.08/s, xAI API $0.08/s, ModelsLab $0.096/s, Runway API $0.10/s, Segmind $0.108/s, Higgsfield API. Apps: Poe $0.096/s, Krea $0.16/s (Basic $9/mo), Pollo AI $0.165/s (Lite $15/mo), OpenArt Starter $14/mo.
	T2V I2V Ref Audio Extend	1080p	—	$0.07/sAlibaba Cloud Model Studio	12	5	Jun 2026	1050
Alibaba ATH's cinematic audio-video model for text, image and reference-to-video. API: Alibaba Cloud Model Studio $0.07/s, WaveSpeedAI $0.07/s, OpenRouter $0.099/s, Kie.ai $0.113/s, AI/ML API $0.14/s, each::labs (Eachlabs) $0.14/s, fal.ai $0.14/s, Higgsfield API $0.14/s, ModelsLab $0.14/s, Replicate $0.14/s, Segmind $0.20/s, Together AI $0.14 per undefined s clip. Apps: Pollo AI $0.12/s (Lite $15/mo), Poe $0.14/s, Krea $0.281/s (Basic $9/mo), Figma Weave (Weavy) Starter $24/mo, OpenArt Starter $14/mo.
	T2V I2V Audio	—	10 s	—	0	0	Aug 2026	1041
114B mixture-of-experts open-weight model generating 10 s clips with synchronized audio; API listed as coming soon.
	T2V I2V F+L Ref Audio Extend Edit	1080p	15 s	$0.08/sKie.ai	13	4	Apr 2026	1035
Wan with native audio across text-, image- and reference-to-video plus a video-edit endpoint; latest snapshot 2026-06-12. API: Kie.ai $0.08/s, AI/ML API $0.10/s, Alibaba Cloud Model Studio $0.10/s, DeepInfra $0.10/s, each::labs (Eachlabs) $0.10/s, fal.ai $0.10/s, Higgsfield API $0.10/s, OpenRouter $0.10/s, Replicate $0.10/s, WaveSpeedAI $0.10/s, Segmind $0.143/s, ModelsLab $0.50/s, Together AI $0.10 per undefined s clip. Apps: Pollo AI $0.06/s (Lite $15/mo), Poe $0.10/s, Freepik (Magnific) Premium €15/mo, OpenArt Starter $14/mo.
	T2V I2V Ref V2V Audio Extend Edit	1080p	—	$0.099/sOpenRouter	13	2	Apr 2026	1033
First HappyHorse release, with text-, image-, reference-to-video and a video-edit endpoint. API: OpenRouter $0.099/s, AI/ML API $0.14/s, Alibaba Cloud Model Studio $0.14/s, each::labs (Eachlabs) $0.14/s, fal.ai $0.14/s, Higgsfield API $0.14/s, Kie.ai $0.14/s, ModelsLab $0.14/s, Replicate $0.14/s, WaveSpeedAI $0.14/s, Runway API $0.15/s, Segmind $0.20/s, Together AI $0.24 per undefined s clip. Apps: Poe $0.14/s, Pollo AI Lite $15/mo.
	T2V I2V F+L Ref V2V Audio Edit	4K	15 s	$0.07/sKie.ai	10	8	Feb 2026	1022
All-in-one multimodal Kling 3.0: reference images, video and voice, video editing and native audio with storyboards. API: Kie.ai $0.07/s, each::labs (Eachlabs) $0.084/s, fal.ai $0.084/s, Higgsfield API $0.084/s, Kling AI API $0.084/s, WaveSpeedAI $0.084/s, PiAPI $0.10/s, Runway API $0.13/s, Replicate $0.168/s, Segmind $0.21/s. Apps: Higgsfield $0.075/s (Starter $15/mo), Adobe Firefly $0.10/s (Firefly Standard $9.99/mo), Krea $0.168/s (Basic $9/mo), Poe $0.17/s, Freepik (Magnific) Premium €15/mo, Kling AI app Standard $8.80/mo, OpenArt Starter $14/mo, Pollo AI Lite $15/mo.
	T2V I2V F+L Ref Audio Camera	4K	15 s	$0.07/sKie.ai	13	14	Feb 2026	1000
Kling's audio-video model with multi-shot storyboards, element consistency and native 4K; the anchor model (Elo 1000) of the Artificial Analysis video arena. API: Kie.ai $0.07/s, AI/ML API $0.084/s, fal.ai $0.084/s, Higgsfield API $0.084/s, Kling AI API $0.084/s, ModelsLab $0.084/s, OpenRouter $0.084/s, WaveSpeedAI $0.084/s, Runway API $0.09/s, PiAPI $0.10/s, Segmind $0.118/s, Replicate $0.168/s, each::labs (Eachlabs). Apps: Higgsfield $0.078/s (Starter $15/mo), Adobe Firefly $0.10/s (Firefly Standard $9.99/mo), Krea $0.168/s (Basic $9/mo), Poe $0.23/s, Figma Weave (Weavy) Starter $24/mo, Freepik (Magnific) €0.0675/s (Premium €15/mo), Hedra Studio, Kaiber Creator $29/mo, Kling AI app Standard $8.80/mo, Leonardo.Ai Essential $12/mo, LTX Studio Standard $35/mo, OpenArt Starter $14/mo, Pollo AI Lite $15/mo, Runway (app) Standard $15/mo.
	T2V I2V Audio	1080p	15 s	$0.02/seach::labs (Eachlabs)	5	0	Aug 2026	990
xAI's lowest-priced video model with lip-synced audio; 1080p is upscaled from 720p. API: each::labs (Eachlabs) $0.02/s, fal.ai $0.02/s, OpenRouter $0.02/s, Runway API $0.02/s, xAI API $0.02/s.
	T2V I2V F+L Ref Audio Extend Camera	1080p	15 s	$0.02/sKie.ai	10	3	Mar 2026	981
PixVerse's main model with video references and extension at low per-second prices. API: Kie.ai $0.02/s, PixVerse Platform API $0.022/s, DeepInfra $0.025/s, each::labs (Eachlabs) $0.025/s, fal.ai $0.025/s, WaveSpeedAI $0.025/s, Segmind $0.031/s, Replicate $0.05/s, Higgsfield API $0.115/s, Together AI $0.09 per undefined s clip. Apps: Pollo AI $0.075/s (Lite $15/mo), Freepik (Magnific) Premium €15/mo, OpenArt Starter $14/mo.
	T2V I2V F+L Ref Audio Extend	4K	8 s	$0.12/sPiAPI	12	16	Nov 2025	961
Google's top Veo tier: 4 to 8 second clips with native speech and sound up to 4K, with first/last frame, reference images and extend. API: PiAPI $0.12/s, AI/ML API $0.20/s, each::labs (Eachlabs) $0.20/s, Gemini Enterprise Agent Platform (Vertex AI) $0.20/s, OpenRouter $0.20/s, Replicate $0.20/s, Runway API $0.20/s, Segmind $0.20/s, WaveSpeedAI $0.20/s, DeepInfra $0.40/s, Kie.ai $1.25 per undefined s clip, Together AI $0.08 per undefined s clip. Apps: Adobe Firefly $0.25/s (Firefly Standard $9.99/mo), Google Flow $0.312/s (Google AI Plus $4.99/mo), Poe $0.404/s, Higgsfield $0.49/s (Plus $49/mo), Pollo AI $0.563/s (Lite $15/mo), CapCut Pro, Flora Starter $18/mo, Freepik (Magnific) €0.39/s (Premium €15/mo), Hedra Studio, Invideo Pro $30/mo, Krea Pro $35/mo, Leonardo.Ai Essential $12/mo, LTX Studio Pro $125/mo, Monica Pro $8.30/mo, OpenArt Starter $14/mo, Runway (app) Standard $15/mo.
	T2V I2V F+L Ref Audio Extend	4K	8 s	$0.06/sPiAPI	12	9	Nov 2025	957
Faster, cheaper Veo 3.1 with the same audio, reference and extend features. API: PiAPI $0.06/s, AI/ML API $0.08/s, Gemini Enterprise Agent Platform (Vertex AI) $0.08/s, OpenRouter $0.08/s, each::labs (Eachlabs) $0.10/s, Replicate $0.10/s, Runway API $0.10/s, Segmind $0.10/s, WaveSpeedAI $0.10/s, DeepInfra $0.15/s, ModelsLab $0.24/s, Kie.ai $0.30 per undefined s clip. Apps: Google Flow $0.062/s (Google AI Plus $4.99/mo), Adobe Firefly $0.10/s (Firefly Standard $9.99/mo), Poe $0.135/s, Krea $0.161/s (Basic $9/mo), Higgsfield $0.196/s (Plus $49/mo), Pollo AI $0.281/s (Lite $15/mo), Freepik (Magnific) Premium €15/mo, Leonardo.Ai Essential $12/mo, LTX Studio Pro $125/mo.
	T2V I2V F+L Audio Extend	1080p	8 s	$0.03/sAI/ML API	9	7	Apr 2026	951
Lowest-cost Veo 3.1 tier with native audio, in preview on Vertex AI. API: AI/ML API $0.03/s, each::labs (Eachlabs) $0.03/s, Gemini Enterprise Agent Platform (Vertex AI) $0.03/s, OpenRouter $0.03/s, Replicate $0.05/s, WaveSpeedAI $0.05/s, ModelsLab $0.096/s, Kie.ai $0.15 per undefined s clip, Together AI $0.05 per undefined s clip. Apps: Google Flow $0.031/s (Google AI Plus $4.99/mo), Adobe Firefly $0.05/s (Firefly Standard $9.99/mo), Poe $0.05/s, Krea $0.06/s (Basic $9/mo), Pollo AI $0.188/s (Lite $15/mo), Leonardo.Ai Essential $12/mo, LTX Studio Pro $125/mo.
	T2V I2V F+L Lips Audio Camera	4K	20 s	$0.03/sReplicate	7	2	Aug 2026	944
Faster, cheaper LTX-2.5 with clips up to 20s at 720p/1080p (24/25 fps); open weights. API: Replicate $0.03/s, each::labs (Eachlabs) $0.09/s, fal.ai $0.09/s, LTX API $0.09/s, WaveSpeedAI $0.10/s, Segmind $0.113/s, Higgsfield API. Apps: Krea $0.18/s (Basic $9/mo), LTX Studio Lite $15/mo.
	T2V I2V F+L Lips Audio Camera	4K	10 s	$0.12/seach::labs (Eachlabs)	6	2	Aug 2026	940
Higher-fidelity LTX-2.5 variant with native multi-shot scenes; audio-to-video billed per input-audio second at the same rates. API: each::labs (Eachlabs) $0.12/s, fal.ai $0.12/s, LTX API $0.12/s, ModelsLab $0.144/s, Segmind $0.15/s, Higgsfield API. Apps: Krea $0.24/s (Basic $9/mo), LTX Studio Lite $15/mo.
	T2V I2V F+L Audio	1080p	16 s	$0.042/sVidu API	5	1	Jan 2026	938
Vidu's quality tier for 1-16 s audio-video clips from text, an image or first/last frames. API: Vidu API $0.042/s, WaveSpeedAI $0.05/s, fal.ai $0.07/s, Replicate $0.07/s, ModelsLab $0.125/s. Apps: Pollo AI $0.15/s (Lite $15/mo).
	T2V I2V F+L Ref Audio	1080p	16 s	$0.033/sVidu API	6	1	Feb 2026	913
Faster, cheaper Vidu Q3, also used for reference-to-video. API: Vidu API $0.033/s, fal.ai $0.035/s, Replicate $0.04/s, WaveSpeedAI $0.04/s, ModelsLab $0.06/s, Together AI $0.20 per undefined s clip. Apps: Pollo AI $0.113/s (Lite $15/mo).
	T2V I2V	720p	10 s	$0.12/sOpenRouter	3	5	Dec 2025	910
Runway's flagship text- and image-to-video model, 2 to 10 second 720p clips with ProRes and native HDR delivery options. API: OpenRouter $0.12/s, Replicate $0.12/s, Runway API $0.12/s. Apps: Adobe Firefly $0.20/s (Firefly Standard $9.99/mo), Krea $0.24/s (Basic $9/mo), Runway app $0.288/s (Standard $15/mo), Poe $0.30/s, Freepik (Magnific) Premium €15/mo.
	T2V I2V F+L Ref V2V Edit	1080p	10 s	$0.078/sPiAPI	11	5	Dec 2025	890
Kling's first unified 'omni' video model: generation and prompt-driven editing from text, images and video in one endpoint. API: PiAPI $0.078/s, fal.ai $0.084/s, Higgsfield API $0.084/s, Kling AI API $0.084/s, Replicate $0.084/s, WaveSpeedAI $0.084/s, AI/ML API $0.112/s, OpenRouter $0.112/s, Segmind $0.14/s, each::labs (Eachlabs), Magnific API (formerly Freepik API). Apps: Higgsfield $0.105/s (Starter $15/mo), Poe $0.11/s, Krea $0.225/s (Basic $9/mo), Freepik (Magnific) Premium €15/mo, Pollo AI Lite $15/mo.
	T2V I2V F+L Lips Audio Extend Edit	4K	10 s	$0.04/sLTX API	4	1	Mar 2026	884
Previous LTX generation and the one with Retake (edit) and Extend on the API ($0.10/s at 1080p); audio-to-video $0.06-$0.34/s. API: LTX API $0.04/s, fal.ai $0.08/s, Replicate $0.08/s, ModelsLab $0.096/s. Apps: LTX Studio Lite $15/mo.
	T2V I2V F+L Audio	4K	20 s	$0.03/sLTX API	4	1	Mar 2026	839
Cheapest LTX API model, clips up to 20s at 720p/1080p. API: LTX API $0.03/s, fal.ai $0.06/s, Replicate $0.06/s, ModelsLab $0.072/s. Apps: LTX Studio Lite $15/mo.
	T2V I2V F+L Loop	1080p	5 s	$0.02/sPika API	1	1	Nov 2025	832
Pika's own text- and image-to-video model, with Pikaframes keyframe transitions (1-5 keyframes, optional loop). API: Pika API $0.02/s. Apps: Pika app $0.044/s (Starter $10/mo).
	T2V Audio	—	—	$0.01/sfal.ai	1	0	Sep 2026	830
Creatify's video model, listed on the Artificial Analysis text-to-video arena. API: fal.ai $0.01/s.
	T2V I2V F+L V2V Motion Extend Edit Camera	1080p	10 s	$0.03/sLuma API	7	1	Jun 2026	820
Luma's production video model: up to 16 keyframes per clip, Modify Video up to 20s, reframe, motion transfer, native 16-bit HDR and EXR export; no native audio. API: Luma API $0.03/s, Replicate $0.03/s, each::labs (Eachlabs) $0.10/s, fal.ai $0.10/s, WaveSpeedAI $0.10/s, Segmind $0.125/s, AI/ML API. Apps: Luma Dream Machine $0.03/s (Plus $30/mo).
	T2V I2V Audio	1080p	20 s	$0.02/sDeepInfra	5	0	Feb 2026	801
Pruna's first fast video model: up to 1080p and 20s at 24 or 48 fps, with draft mode. API: DeepInfra $0.02/s, each::labs (Eachlabs) $0.02/s, Pruna API $0.02/s, Replicate $0.02/s, WaveSpeedAI $0.02/s.
	T2V I2V F+L	1080p	10 s	$0.04/sPiAPI	13	7	Oct 2025	—
Cheapest current Kling for silent clips, with first/last frame control in Pro mode. API: PiAPI $0.04/s, fal.ai $0.042/s, Higgsfield API $0.042/s, Kie.ai $0.042/s, Kling AI API $0.042/s, ModelsLab $0.042/s, WaveSpeedAI $0.042/s, AI/ML API $0.07/s, Replicate $0.07/s, Segmind $0.088/s, Runway API $0.12/s, each::labs (Eachlabs), Magnific API (formerly Freepik API). Apps: Poe $0.042/s, Higgsfield $0.06/s (Starter $15/mo), Adobe Firefly $0.075/s (Firefly Standard $9.99/mo), Krea $0.084/s (Basic $9/mo), Freepik (Magnific) €0.021/s (Premium €15/mo), Kaiber Creator $29/mo, Pollo AI Lite $15/mo.
	T2V I2V Ref V2V Audio Extend	1080p	15 s	$0.04/sOpenRouter	14	6	Dec 2025	—
Wan with native audio and multi-shot storytelling, 2-15 s. API: OpenRouter $0.04/s, Kie.ai $0.07/s, PiAPI $0.08/s, AI/ML API $0.10/s, Alibaba Cloud Model Studio $0.10/s, DeepInfra $0.10/s, each::labs (Eachlabs) $0.10/s, fal.ai $0.10/s, Higgsfield API $0.10/s, Replicate $0.10/s, WaveSpeedAI $0.10/s, Segmind $0.125/s, ModelsLab $0.50/s, Magnific API (formerly Freepik API). Apps: Higgsfield $0.075/s (Starter $15/mo), Poe $0.09/s, Krea $0.20/s (Basic $9/mo), Freepik (Magnific) Premium €15/mo, Leonardo.Ai Essential $12/mo, Pollo AI Lite $15/mo.
	T2V I2V F+L Ref V2V Audio Extend Edit	4K	15 s	$0.04/sOpenRouter	11	8	Feb 2026	—
Faster, cheaper Seedance 2.0 with the same feature set, capped at 720p. API: OpenRouter $0.04/s, PiAPI $0.048/s, BytePlus ModelArk $0.056/s, Segmind $0.056/s, AI/ML API $0.057/s, Kie.ai $0.059/s, Replicate $0.07/s, ModelsLab $0.08/s, WaveSpeedAI $0.10/s, fal.ai $0.108/s, Runway API $0.11/s. Apps: Poe $0.122/s, Krea $0.129/s (Basic $9/mo), Adobe Firefly $0.15/s (Firefly Standard $9.99/mo), Higgsfield $0.18/s (Starter $15/mo), Pollo AI $0.24/s (Lite $15/mo), Freepik (Magnific) €0.1762/s (Premium €15/mo), Invideo Basic $9/mo, Kaiber Creator $29/mo.
	T2V I2V Audio	1080p	10 s	$0.04/sPiAPI	11	8	Dec 2025	—
First Kling with native audio ('see the sound, hear the visual'), including custom voice reference. API: PiAPI $0.04/s, Kling AI API $0.042/s, WaveSpeedAI $0.042/s, Kie.ai $0.055/s, AI/ML API $0.07/s, each::labs (Eachlabs) $0.07/s, fal.ai $0.07/s, Higgsfield API $0.07/s, Replicate $0.07/s, Segmind $0.087/s, Magnific API (formerly Freepik API). Apps: Poe $0.071/s, Higgsfield $0.075/s (Starter $15/mo), Krea $0.14/s (Basic $9/mo), Freepik (Magnific) Premium €15/mo, Kaiber Creator $29/mo, Kling AI app Standard $8.80/mo, LTX Studio Standard $35/mo, Pollo AI Lite $15/mo.
	T2V I2V F+L Ref V2V Audio Extend Edit Camera	1080p	30 s	$0.142/sKie.ai	12	7	Oct 2026	—
Kuaishou's flagship: native 30-second clips with up to 10 keyframes, omni reference (images, video, voice) and video editing in one model. API: Kie.ai $0.142/s, Kling AI API $0.154/s, AI/ML API $0.157/s, each::labs (Eachlabs) $0.157/s, fal.ai $0.157/s, Higgsfield API $0.157/s, Replicate $0.157/s, Runway API $0.157/s, WaveSpeedAI $0.157/s, Segmind $0.166/s, PiAPI $0.167/s, ModelsLab. Apps: Adobe Firefly $0.175/s (Firefly Standard $9.99/mo), OpenArt $0.28/s (Pro $56/mo), Krea $0.316/s (Basic $9/mo), Runway (app) $0.566/s (Standard $15/mo), Freepik (Magnific) €0.21/s (Premium €15/mo), Kling AI app Standard $8.80/mo, Pollo AI Lite $15/mo.
	T2V I2V	1080p	10 s	$0.025/sKie.ai	13	6	Oct 2025	—
2025 Hailuo with stronger body movement, facial expression and physics; 1080p at 6 s. API: Kie.ai $0.025/s, AI/ML API $0.028/s, PiAPI $0.038/s, each::labs (Eachlabs) $0.047/s, fal.ai $0.047/s, Higgsfield API $0.047/s, MiniMax API $0.047/s, Replicate $0.047/s, WaveSpeedAI $0.047/s, Segmind $0.058/s, ModelsLab $0.067/s, OpenRouter $0.082/s, Magnific API (formerly Freepik API). Apps: Higgsfield $0.075/s (Starter $15/mo), Krea $0.094/s (Basic $9/mo), Freepik (Magnific) Premium €15/mo, Hedra Studio, Leonardo.Ai Essential $12/mo, Pollo AI Lite $15/mo.
	T2V I2V F+L Ref V2V Audio Extend Edit	4K	15 s	$0.019/sKie.ai	11	6	Jun 2026	—
Lowest-cost Seedance 2.0 tier for volume drafts, up to 720p. API: Kie.ai $0.019/s, OpenRouter $0.034/s, Segmind $0.035/s, BytePlus ModelArk $0.036/s, Replicate $0.04/s, PiAPI $0.042/s, ModelsLab $0.044/s, WaveSpeedAI $0.06/s, Runway API $0.07/s, fal.ai $0.072/s, AI/ML API. Apps: CapCut $0.007/s (Pro), Poe $0.07/s, Krea $0.082/s (Basic $9/mo), Pollo AI $0.15/s (Lite $15/mo), Higgsfield Starter $15/mo, Invideo Pro $30/mo.
	Ref Motion	1080p	—	$0.07/sReplicate	11	5	Mar 2026	—
Copies the motion and expressions of a reference video onto a character image, with element binding for consistency. API: Replicate $0.07/s, Kie.ai $0.10/s, AI/ML API $0.126/s, fal.ai $0.126/s, Higgsfield API $0.126/s, Kling AI API $0.126/s, ModelsLab $0.126/s, WaveSpeedAI $0.126/s, Segmind $0.148/s, PiAPI $0.195/s, each::labs (Eachlabs). Apps: Higgsfield $0.069/s (Plus $49/mo), Poe $0.14/s, Freepik (Magnific) Premium €15/mo, Kling AI app Standard $8.80/mo, OpenArt Starter $14/mo.
	T2V I2V F+L	1080p	10 s	$0.017/sMiniMax API	11	5	Jun 2025	—
MiniMax's mid-2025 model that brought 1080p and 10 s clips; still the cheapest Hailuo at 512P. API: MiniMax API $0.017/s, Replicate $0.017/s, WaveSpeedAI $0.017/s, fal.ai $0.017/s, Segmind $0.021/s, Kie.ai $0.025/s, AI/ML API $0.028/s, each::labs (Eachlabs) $0.045/s, Together AI $0.049/s, ModelsLab $0.096/s, Magnific API (formerly Freepik API). Apps: Krea $0.033/s (Basic $9/mo), Higgsfield $0.037/s (Starter $15/mo), Poe $0.045/s, Freepik (Magnific) Premium €15/mo, Pollo AI Lite $15/mo.
	T2V I2V Ref V2V Audio Extend Edit Upscale	1080p	15 s	$0.012/sKie.ai	10	5	—	—
xAI's original video model, now the one that edits and extends existing videos (edits capped at 8.7s and 720p). API: Kie.ai $0.012/s, AI/ML API $0.05/s, each::labs (Eachlabs) $0.05/s, fal.ai $0.05/s, OpenRouter $0.05/s, Replicate $0.05/s, WaveSpeedAI $0.05/s, xAI API $0.05/s, ModelsLab $0.06/s, Segmind $0.068/s. Apps: Poe $0.051/s, Krea $0.10/s (Basic $9/mo), Higgsfield $0.105/s (Starter $15/mo), Freepik (Magnific) Premium €15/mo, Pollo AI Lite $15/mo.
	Motion	1080p	30 s	$0.055/sKie.ai	12	2	Dec 2025	—
Motion transfer from a reference video (up to 30 s) onto a character image at the lowest Kling motion-control rate. API: Kie.ai $0.055/s, PiAPI $0.065/s, AI/ML API $0.07/s, fal.ai $0.07/s, Higgsfield API $0.07/s, Kling AI API $0.07/s, ModelsLab $0.07/s, Replicate $0.07/s, WaveSpeedAI $0.07/s, Segmind $0.087/s, each::labs (Eachlabs), Magnific API (formerly Freepik API). Apps: Higgsfield $0.075/s (Starter $15/mo), Freepik (Magnific) Premium €15/mo.
	I2V	1080p	10 s	$0.019/sAI/ML API	9	5	Oct 2025	—
Faster, cheaper Hailuo 2.3 for image-to-video. API: AI/ML API $0.019/s, PiAPI $0.027/s, each::labs (Eachlabs) $0.032/s, fal.ai $0.032/s, MiniMax API $0.032/s, Replicate $0.032/s, WaveSpeedAI $0.032/s, ModelsLab $0.038/s, Segmind $0.04/s. Apps: Higgsfield $0.05/s (Starter $15/mo), Krea $0.063/s (Basic $9/mo), Freepik (Magnific) Premium €15/mo, Leonardo.Ai Essential $12/mo, Pollo AI Lite $15/mo.
	T2V I2V F+L	1080p	12 s	$0.014/sKie.ai	9	5	Jun 2025	—
ByteDance's 2025 silent 1080p model, still sold on ModelArk at a fraction of Seedance 2.x prices. API: Kie.ai $0.014/s, BytePlus ModelArk $0.024/s, fal.ai $0.024/s, Replicate $0.03/s, WaveSpeedAI $0.03/s, Together AI $0.114/s, AI/ML API, each::labs (Eachlabs), Segmind. Apps: Higgsfield $0.06/s (Starter $15/mo), Krea $0.112/s (Basic $9/mo), Freepik (Magnific) Premium €15/mo, Poe, Pollo AI Lite $15/mo.
	T2V I2V Lips Audio Extend	1080p	10 s	$0.05/sAI/ML API	10	4	Sep 2025	—
First Wan with audio-synced generation, 5 or 10 s clips. API: AI/ML API $0.05/s, Alibaba Cloud Model Studio $0.05/s, each::labs (Eachlabs) $0.05/s, fal.ai $0.05/s, Replicate $0.05/s, WaveSpeedAI $0.05/s, Kie.ai $0.06/s, Segmind $0.063/s, ModelsLab $0.50/s, Magnific API (formerly Freepik API). Apps: Higgsfield $0.075/s (Starter $15/mo), Krea $0.10/s (Basic $9/mo), Freepik (Magnific) Premium €15/mo, Pollo AI Lite $15/mo.
	T2V I2V	1080p	12 s	$0.0096/sfal.ai	9	4	Oct 2025	—
Cheapest Seedance on ModelArk: silent 1080p text- and image-to-video. API: fal.ai $0.0096/s, BytePlus ModelArk $0.01/s, WaveSpeedAI $0.012/s, Replicate $0.015/s, Kie.ai $0.016/s, ModelsLab $0.024/s, AI/ML API, each::labs (Eachlabs), Segmind. Apps: Krea $0.021/s (Basic $9/mo), Higgsfield $0.06/s (Starter $15/mo), Freepik (Magnific) Premium €15/mo, Pollo AI Lite $15/mo.
	T2V I2V F+L V2V	1080p	—	$0.02/seach::labs (Eachlabs)	10	3	Jul 2025	—
Apache-2.0 mixture-of-experts text- and image-to-video weights you can self-host. API: each::labs (Eachlabs) $0.02/s, WaveSpeedAI $0.03/s, DeepInfra $0.033/s, fal.ai $0.04/s, Kie.ai $0.04/s, SiliconFlow $0.058/s, ModelsLab $0.075 per undefined s clip, PiAPI $0.28 per undefined s clip, Replicate $0.05 per undefined s clip, Segmind $0.063 per undefined s clip. Apps: Higgsfield $0.075/s (Starter $15/mo), Freepik (Magnific) Premium €15/mo, Krea Basic $9/mo.
	T2V I2V F+L Ref V2V Audio Extend Edit	1080p	30 s	$0.061/sKie.ai	10	3	Aug 2026	—
High-speed Wan 3.0 with the same capabilities and much faster end-to-end generation. API: Kie.ai $0.061/s, Alibaba Cloud Model Studio $0.068/s, each::labs (Eachlabs) $0.068/s, fal.ai $0.068/s, Higgsfield API $0.068/s, OpenRouter $0.068/s, Replicate $0.068/s, Runway API $0.068/s, Segmind $0.072/s, WaveSpeedAI $0.075/s. Apps: Higgsfield $0.113/s (Starter $15/mo), Krea $0.136/s (Basic $9/mo), Pollo AI $0.188/s (Lite $15/mo).
	T2V I2V Audio	1080p	15 s	$0.09/sKie.ai	8	3	Jun 2026	—
Value tier of Kling 3.0 with native audio, text-to-video and image-to-video at 720p or 1080p. API: Kie.ai $0.09/s, AI/ML API $0.112/s, each::labs (Eachlabs) $0.112/s, fal.ai $0.112/s, Higgsfield API $0.112/s, Kling AI API $0.112/s, WaveSpeedAI $0.112/s, PiAPI $0.13/s. Apps: Poe $0.18/s, Krea $0.225/s (Basic $9/mo), Pollo AI Lite $15/mo.
	Lips	1080p	300 s	$0.04/sKie.ai	9	1	Oct 2025	—
Talking avatar from one image plus audio or text, with natural expressions; videos up to 5 minutes. API: Kie.ai $0.04/s, WaveSpeedAI $0.05/s, PiAPI $0.052/s, AI/ML API $0.056/s, Kling AI API $0.056/s, Replicate $0.056/s, fal.ai $0.056/s, Segmind $0.07/s, each::labs (Eachlabs). Apps: Poe $0.12/s.
	I2V Lips	1080p	—	$0.12/seach::labs (Eachlabs)	9	1	—	—
Sharper OmniHuman with more expressive motion; image plus audio to talking video at up to 1080p. API: each::labs (Eachlabs) $0.12/s, PiAPI $0.13/s, Kie.ai $0.135/s, ModelsLab $0.14/s, AI/ML API $0.16/s, fal.ai $0.16/s, Replicate $0.16/s, WaveSpeedAI $0.16/s, Magnific API (formerly Freepik API). Apps: Freepik (Magnific) Premium €15/mo.
	T2V I2V F+L V2V Extend	1080p	8 s	$0.03/sfal.ai	8	2	Aug 2025	—
PixVerse V5 with multi-transition videos. API: fal.ai $0.03/s, PixVerse Platform API $0.04/s, WaveSpeedAI $0.045/s, Replicate $0.06/s, Together AI $0.06/s, Segmind $0.075/s, AI/ML API, Magnific API (formerly Freepik API). Apps: Freepik (Magnific) Premium €15/mo, Pollo AI Lite $15/mo.
	I2V Ref V2V Audio	1080p	—	$0.02/sKie.ai	9	1	—	—
Budget Wan 2.6 for image- and reference-to-video, with optional audio. API: Kie.ai $0.02/s, AI/ML API $0.025/s, Alibaba Cloud Model Studio $0.025/s, WaveSpeedAI $0.025/s, Segmind $0.031/s, each::labs (Eachlabs) $0.05/s, fal.ai $0.05/s, Replicate $0.05/s, ModelsLab $0.10/s. Apps: Pollo AI Lite $15/mo.
	I2V	720p	10 s	$0.05/sAI/ML API	8	1	—	—
Runway's fastest and cheapest image-to-video model. API: AI/ML API $0.05/s, each::labs (Eachlabs) $0.05/s, Replicate $0.05/s, Runway API $0.05/s, WaveSpeedAI $0.056/s, ModelsLab $0.08/s, Segmind $0.10/s, Magnific API (formerly Freepik API). Apps: Pollo AI $0.30/s (Lite $15/mo).
	T2V I2V V2V Lips Audio Edit	4K	—	$0.06/sAI/ML API	6	2	—	—
Lightricks' hosted LTX-2 quality tier: synced audio and video up to 4K, plus retake (edit a section) and audio-to-video. API: AI/ML API $0.06/s, Replicate $0.06/s, WaveSpeedAI $0.06/s, ModelsLab $0.07/s, Segmind $0.10/s, Magnific API (formerly Freepik API). Apps: Poe $0.061/s, Freepik (Magnific) Premium €15/mo.
	T2V I2V F+L Ref Audio	1080p	10 s	$0.031/sPixVerse Platform API	6	2	Jan 2026	—
Cost-efficient PixVerse for fixed 5/8/10 s clips. API: PixVerse Platform API $0.031/s, each::labs (Eachlabs) $0.044/s, fal.ai $0.07/s, Replicate $0.07/s, WaveSpeedAI $0.07/s, Together AI $0.13 per undefined s clip. Apps: Poe $0.07/s, Pollo AI Lite $15/mo.
	T2V I2V F+L Ref V2V Edit	720p	—	$0.04/sWaveSpeedAI	6	2	Feb 2025	—
Apache-2.0 Wan 2.1 family (T2V/I2V 14B, 1.3B, VACE, FLF2V) that seeded much of the open video ecosystem. API: WaveSpeedAI $0.04/s, Replicate $0.07/s, ModelsLab $0.075/s, fal.ai $0.08/s, DeepInfra $0.40 per undefined s clip, PiAPI $0.28 per undefined s clip. Apps: Krea Basic $9/mo, Pollo AI Lite $15/mo.
	T2V I2V V2V Lips Audio Extend	—	—	$0.02/sWaveSpeedAI	4	3	Mar 2026	—
The open 22B LTX-2.3 checkpoint run by third-party hosts with LoRA support; audio and video in one pass. API: WaveSpeedAI $0.02/s, each::labs (Eachlabs) $0.025/s, fal.ai $0.035/s, ModelsLab $0.075 per undefined s clip. Apps: Krea $0.072/s (Basic $9/mo), Figma Weave (Weavy) Starter $24/mo, OpenArt Starter $14/mo.
	T2V I2V F+L Ref Audio Edit	768p	15 s	$0.03/sReplicate	6	1	Jul 2026	—
H3 tuned for speed and everyday content at 480P or 768P. API: Replicate $0.03/s, AI/ML API $0.04/s, MiniMax API $0.04/s, Segmind $0.042/s, each::labs (Eachlabs) $0.06/s, WaveSpeedAI $0.06/s. Apps: Pollo AI Lite $15/mo.
	T2V I2V F+L Ref Audio Edit	768p	15 s	$0.013/sReplicate	6	1	Jul 2026	—
Lowest-cost H3 for creative exploration and batch generation. API: Replicate $0.013/s, AI/ML API $0.025/s, each::labs (Eachlabs) $0.025/s, MiniMax API $0.025/s, WaveSpeedAI $0.025/s, Segmind $0.026/s. Apps: Pollo AI Lite $15/mo.
	T2V I2V Ref	720p	6 s	$0.083/sfal.ai	5	2	—	—
The original Hailuo model: 6 s 720p clips, with a subject-reference variant for consistent characters. API: fal.ai $0.083/s, Replicate $0.083/s, AI/ML API $0.43 per undefined s clip, Segmind $0.63 per undefined s clip, WaveSpeedAI $0.50 per undefined s clip. Apps: Krea $0.173/s (Basic $9/mo), Pollo AI Lite $15/mo.
	Lips	—	—	$0.12/seach::labs (Eachlabs)	6	1	—	—
Talking or singing video from one photo of a person plus an audio track, with body motion that follows the audio. API: each::labs (Eachlabs) $0.12/s, WaveSpeedAI $0.12/s, AI/ML API $0.14/s, fal.ai $0.14/s, Replicate $0.14/s, ModelsLab $0.168/s. Apps: Poe $0.14/s.
	T2V I2V F+L Audio	1080p	10 s	$0.03/sfal.ai	5	2	Dec 2025	—
PixVerse with single- and multi-shot modes and optional audio. API: fal.ai $0.03/s, PixVerse Platform API $0.04/s, each::labs (Eachlabs) $0.056/s, WaveSpeedAI $0.09/s, AI/ML API. Apps: Pollo AI $0.113/s (Lite $15/mo), Freepik (Magnific) Premium €15/mo.
	V2V Edit	—	30 s	$0.28/sOpenRouter	4	3	Jun 2026	—
In-context video editing: change an existing 2 to 30 second clip by prompt, with up to 5 keyframe images. API: OpenRouter $0.28/s, Runway API $0.28/s, Replicate $0.336/s, WaveSpeedAI $0.37/s. Apps: Adobe Firefly $0.499/s (Firefly Standard $9.99/mo), Runway app $0.672/s (Standard $15/mo), Figma Weave (Weavy) Starter $24/mo.
	I2V Lips	720p	300 s	$0.07/sWaveSpeedAI	6	1	Sep 2025	—
Turns any image plus audio or text into a talking video up to 5 minutes long. API: WaveSpeedAI $0.07/s, AI/ML API $0.08/s, each::labs (Eachlabs) $0.08/s, fal.ai $0.08/s, Replicate $0.08/s, Segmind $0.084/s. Apps: Freepik (Magnific) Premium €15/mo.
	Motion Swap	—	—	$0.02/sReplicate	5	2	Sep 2025	—
Animates a character image with a driving video (Move) or swaps the person in a video (Mix); weights are open. API: Replicate $0.02/s, Kie.ai $0.03/s, fal.ai $0.04/s, each::labs (Eachlabs) $0.06/s, Alibaba Cloud Model Studio $0.12/s. Apps: Poe $0.076/s, Freepik (Magnific) Premium €15/mo.
	T2V I2V Ref V2V Audio Edit	720p	20 s	$0.092/sKling AI API	5	1	Nov 2026	—
Faster, cheaper Kling 4.0 for high-volume drafts: 3-20 second clips at 720p with omni reference and editing. API: Kling AI API $0.092/s, fal.ai $0.094/s, PiAPI $0.10/s, each::labs (Eachlabs), Higgsfield API. Apps: Krea $0.189/s (Basic $9/mo).
	Lips	—	60 s	$0.014/sfal.ai	5	1	—	—
Re-syncs the lips in an existing video to new audio or TTS, including multi-person scenes; clips up to 60 s. API: fal.ai $0.014/s, Kling AI API $0.014/s, Replicate $0.014/s, PiAPI $0.02/s, WaveSpeedAI $0.03/s. Apps: Higgsfield $0.02/s (Plus $49/mo).
	T2V I2V Audio	4K	20 s	$0.04/sAI/ML API	4	2	—	—
Lightricks' hosted LTX-2 speed tier for cheap drafts with audio, clips up to 20 s. API: AI/ML API $0.04/s, Replicate $0.04/s, WaveSpeedAI $0.04/s, Segmind $0.075/s. Apps: Poe $0.04/s, Freepik (Magnific) Premium €15/mo.
	T2V I2V F+L V2V Extend Edit Loop Camera	4K	—	$0.08/sWaveSpeedAI	4	2	—	—
Luma's 2025 Ray2, still served on Luma's legacy Dream Machine API (ray-2) and through AWS Bedrock, Adobe Firefly and API resellers; Luma's current API centres on Ray3.2. API: WaveSpeedAI $0.08/s, AWS Bedrock $0.75/s, AI/ML API, Replicate. Apps: Adobe Firefly $0.10/s (Firefly Standard $9.99/mo), Krea $0.16/s (Basic $9/mo).
	T2V I2V	720p	6 s	$0.083/sfal.ai	3	3	—	—
Video-01 tuned for animating illustrations and Live2D-style art. API: fal.ai $0.083/s, Magnific API (formerly Freepik API), Replicate $0.50 per undefined s clip. Apps: Krea $0.173/s (Basic $9/mo), Freepik (Magnific) Premium €15/mo, Pollo AI Lite $15/mo.
	T2V I2V Audio	1080p	15 s	$0.022/seach::labs (Eachlabs)	6	0	Oct 2026	—
PixVerse's highest-quality tier for text- and image-to-video with optional native audio. API: each::labs (Eachlabs) $0.022/s, Kie.ai $0.024/s, PixVerse Platform API $0.027/s, WaveSpeedAI $0.027/s, Segmind $0.032/s, Replicate $0.06/s.
	T2V I2V F+L Ref	1080p	8 s	$0.023/sVidu API	4	2	—	—
Earlier Vidu for quick 4-8 s image-to-video and reference clips. API: Vidu API $0.023/s, each::labs (Eachlabs) $0.05/s, fal.ai $0.05/s, WaveSpeedAI $0.06/s. Apps: Poe $0.20 per undefined s clip, Pollo AI Lite $15/mo.
	I2V Ref Audio	4K	16 s	$0.032/sfal.ai	5	1	—	—
Preview of Shengshu's next Vidu for reference- and image-to-video up to 4K. API: fal.ai $0.032/s, Kie.ai $0.036/s, Vidu API $0.042/s, each::labs (Eachlabs) $0.045/s, WaveSpeedAI $0.045/s. Apps: Pollo AI Lite $15/mo.
	T2V I2V Ref Audio Extend	1080p	15 s	$0.025/sfal.ai	2	3	Sep 2026	—
fal's post-trained, speed-optimized variant of MiniMax H3 for cheap 1-15 s clips with audio. API: fal.ai $0.025/s, Segmind $0.026/s. Apps: Krea $0.05/s (Basic $9/mo), Freepik (Magnific) €0.03/s (Premium €15/mo), Pollo AI Lite $15/mo.
	I2V Lips	1080p	—	$0.045/sWaveSpeedAI	5	0	—	—
Turns one portrait photo and a voice track into an expressive talking-head video. API: WaveSpeedAI $0.045/s, OpenRouter $0.05/s, Replicate $0.075/s, fal.ai $0.10/s, Segmind $0.125/s.
	Lips	4K	—	$0.075/sReplicate	5	0	Apr 2026	—
HeyGen's digital-twin avatar built from a 15 s webcam recording; lip-synced presenter videos up to 4K. API: Replicate $0.075/s, each::labs (Eachlabs) $0.10/s, fal.ai $0.10/s, Segmind $0.10/s, WaveSpeedAI $0.14/s.
	Lips Camera	720p	300 s	$0.07/sBytePlus	5	0	—	—
Talking and singing avatar from one image plus audio, with camera and body movement, multi-person scenes and non-human subjects; up to 300 s. API: BytePlus $0.07/s, Kie.ai $0.07/s, Segmind $0.087/s, WaveSpeedAI $0.10/s, fal.ai $0.105/s.
	Lips	—	—	$0.02/seach::labs (Eachlabs)	5	0	—	—
Lip-syncs an existing video to uploaded audio or to text-to-speech. API: each::labs (Eachlabs) $0.02/s, Segmind $0.03/s, fal.ai $0.04/s, Replicate $0.04/s, WaveSpeedAI $0.04/s.
	Motion	—	—	$0.05/sAI/ML API	4	1	Jul 2025	—
Runway's motion-capture model: drives a character image or video with a performance video. API: AI/ML API $0.05/s, each::labs (Eachlabs) $0.05/s, Runway API $0.05/s, Magnific API (formerly Freepik API). Apps: Freepik (Magnific) Premium €15/mo.
	T2V I2V F+L Ref	1080p	5 s	$0.044/sTogether AI	4	1	—	—
Vidu's 2025 1080p model (plus a 'classic' variant), 5 s clips including anime style. API: Together AI $0.044/s, Vidu API $0.075/s, fal.ai $0.08/s, WaveSpeedAI $0.40 per undefined s clip. Apps: Pollo AI Lite $15/mo.
	T2V I2V F+L Ref Extend	1080p	—	$0.033/sVidu API	3	2	—	—
Q2 text-to-video and reference-to-video. API: Vidu API $0.033/s, WaveSpeedAI $0.035/s, fal.ai $0.075/s. Apps: Krea $0.60/s (Basic $9/mo), Pollo AI Lite $15/mo.
	T2V I2V F+L Ref Audio	1080p	16 s	$0.033/sVidu API	3	2	—	—
Standard Q3 reference-to-video tier. API: Vidu API $0.033/s, WaveSpeedAI $0.07/s, Together AI $0.098 per undefined s clip. Apps: Krea $0.14/s (Basic $9/mo), Pollo AI Lite $15/mo.
	Motion Swap	—	—	$0.05/seach::labs (Eachlabs)	4	0	—	—
Moves the characters in a still image with the motion of a driving video; handles non-human and multi-character scenes. API: each::labs (Eachlabs) $0.05/s, fal.ai $0.05/s, Replicate $0.05/s, WaveSpeedAI $0.05/s.
	T2V	720p	—	$0.08/sfal.ai	4	0	Dec 2024	—
Tencent's original 13B open text-to-video model. API: fal.ai $0.08/s, WaveSpeedAI $0.08/s, PiAPI $0.09 per undefined s clip, Replicate $2.55 per undefined s clip.
	Lips	720p	—	$0.015/sKie.ai	3	1	Aug 2025	—
Open audio-driven dubbing model for unlimited-length talking video from a photo or an existing video. API: Kie.ai $0.015/s, each::labs (Eachlabs) $0.03/s, WaveSpeedAI $0.03/s. Apps: Higgsfield $0.176/s (Plus $49/mo).
	T2V I2V Motion Lips Audio	—	—	$0.012/sWaveSpeedAI	3	1	Jan 2026	—
Lightricks' first open-weight audio+video model (19B); no longer on the LTX API, which now serves LTX-2.3 and LTX-2.5. API: WaveSpeedAI $0.012/s, Replicate $0.02/s, DeepInfra $0.036/s. Apps: Krea $0.018/s (Basic $9/mo).
	T2V I2V Audio	1080p	20 s	$0.025/seach::labs (Eachlabs)	4	0	—	—
Pruna's second-generation P-Video: text, image and audio-conditioned video up to 1080p and 20s at 24 or 48 fps. API: each::labs (Eachlabs) $0.025/s, Pruna API $0.025/s, Segmind $0.025/s, WaveSpeedAI $0.025/s.
	T2V I2V F+L Ref Audio	1080p	15 s	$0.027/sPixVerse Platform API	4	0	Apr 2026	—
PixVerse's cinema-oriented model for complex motion and visual effects. API: PixVerse Platform API $0.027/s, each::labs (Eachlabs) $0.03/s, fal.ai $0.03/s, WaveSpeedAI $0.03/s.
	Motion	720p	—	$0.04/sPixVerse Platform API	4	0	Feb 2026	—
Animates a character with the motion, expressions, speech or singing of a reference video. API: PixVerse Platform API $0.04/s, each::labs (Eachlabs) $0.045/s, Segmind $0.056/s, WaveSpeedAI $0.09/s.
	T2V I2V F+L	1080p	8 s	$0.03/sfal.ai	3	1	—	—
PixVerse's V4 model, 5 or 8 s clips up to 1080p, with a fast mode. API: fal.ai $0.03/s, each::labs (Eachlabs) $0.056/s, Replicate. Apps: Pollo AI Lite $15/mo.
	I2V	720p	10 s	—	0	4	—	—
Runway's image-to-video model with consistent characters across shots from one reference; in the Runway app only (the API sells Gen-4 Turbo and Gen-4.5). Apps: Adobe Firefly $0.10/s (Firefly Standard $9.99/mo), Krea $0.10/s (Basic $9/mo), Runway app $0.288/s (Standard $15/mo), Freepik (Magnific) Premium €15/mo.
	Lips	—	—	$0.05/sfal.ai	4	0	Aug 2025	—
Zero-shot lip sync that matches any speaker's mouth to new audio while keeping their style. API: fal.ai $0.05/s, Replicate $0.05/s, WaveSpeedAI $0.05/s, ModelsLab $0.07/s.
	Lips	—	—	$0.083/sReplicate	4	0	—	—
Higher-fidelity lipsync-2 that keeps teeth, beards and fine facial detail. API: Replicate $0.083/s, fal.ai $0.083/s, Segmind $0.084/s, WaveSpeedAI $0.084/s.
	Lips Edit	—	15 s	$0.167/sfal.ai	4	0	—	—
Re-performs a clip to new audio: lip sync plus matching expressions and head motion, up to 15 s. API: fal.ai $0.167/s, Replicate $0.167/s, WaveSpeedAI $0.167/s, Segmind $0.208/s.
	Lips	—	—	$0.07/seach::labs (Eachlabs)	4	0	—	—
VEED's production lip-sync model for dubbing an existing video to new audio. API: each::labs (Eachlabs) $0.07/s, fal.ai $0.07/s, WaveSpeedAI $0.075/s, Segmind $0.10/s.
	Lips	720p	—	$0.02/sReplicate	4	0	Aug 2025	—
Audio-driven talking/singing video from one image; open 14B weights. API: Replicate $0.02/s, Kie.ai $0.06/s, Alibaba Cloud Model Studio $0.072/s, fal.ai $0.10/s.
	T2V I2V	720p	—	$0.05/sDeepInfra	3	0	May 2026	—
NVIDIA's open omnimodal world model for physical-AI simulation; makes up to 189-frame video clips from text or a first frame. API: DeepInfra $0.05/s, fal.ai $0.05/s, WaveSpeedAI $0.05/s.
	Lips	—	—	$0.067/sHeyGen API	3	0	Apr 2026	—
Dubs an existing video to new audio with avatar-level lip sync; keeps the source resolution. API: HeyGen API $0.067/s, Replicate $0.067/s, fal.ai $0.10/s.
	Lips	—	—	$0.033/sHeyGen API	3	0	Apr 2026	—
Faster, cheaper audio-only lip sync for dubbing an existing video. API: HeyGen API $0.033/s, Replicate $0.033/s, fal.ai $0.05/s.
	I2V Camera	—	—	—	2	1	—	—
Higgsfield's image-to-video model built around cinematic camera-move presets, in Lite, Turbo and Preview tiers. API: each::labs (Eachlabs) $0.13 per undefined s clip, Segmind $0.16 per undefined s clip. Apps: Higgsfield $0.075/s (Starter $15/mo).
	Lips	—	—	$0.005/sfal.ai	3	0	—	—
ByteDance's open lip-sync model that redraws mouths in an existing video to match new audio. API: fal.ai $0.005/s, WaveSpeedAI $0.03/s, Replicate $0.097 per undefined s clip.
	T2V I2V Camera	720p	5 s	$0.06/sReplicate	2	1	—	—
Leonardo's own video model for 5 s text- or image-to-video clips with motion controls. API: Replicate $0.06/s, WaveSpeedAI $0.30 per undefined s clip. Apps: Leonardo.Ai Essential $12/mo.
	T2V I2V Extend	—	—	$0.02/sfal.ai	3	0	Oct 2024	—
Lightricks' original open-weight video model family (0.9.1 to 0.9.8, 2B and 13B), run locally or in ComfyUI. API: fal.ai $0.02/s, Replicate $0.069 per undefined s clip, WaveSpeedAI $0.30 per undefined s clip.
	T2V I2V F+L V2V Extend Loop Camera	4K	—	$0.04/sWaveSpeedAI	2	1	—	—
Faster, cheaper Ray2 Flash, still served on Luma's legacy Dream Machine API (ray-flash-2) and through resellers; Luma's current API centres on Ray3.2. API: WaveSpeedAI $0.04/s, AI/ML API. Apps: Adobe Firefly $0.05/s (Firefly Standard $9.99/mo).
	T2V I2V Audio	—	10 s	$0.04/sfal.ai	3	0	Oct 2025	—
Character.AI's open model that generates video and matching speech or sound in one pass. API: fal.ai $0.04/s, Replicate $0.20 per undefined s clip, WaveSpeedAI $0.15 per undefined s clip.
	I2V Lips Audio	1080p	—	$0.025/sDeepInfra	3	0	—	—
Talking-head video from a single image with a script or audio file, multiple voices and languages. API: DeepInfra $0.025/s, Pruna API $0.025/s, Replicate $0.025/s.
	T2V I2V	1080p	10 s	$0.04/seach::labs (Eachlabs)	2	1	—	—
Pika's 2.2 model: 5 or 10 s text- and image-to-video clips up to 1080p. API: each::labs (Eachlabs) $0.04/s, WaveSpeedAI $0.04/s. Apps: Pollo AI $0.225/s (Lite $15/mo).
	Swap	720p	—	$0.03/sfal.ai	3	0	—	—
Replaces a subject or region in an existing video. API: fal.ai $0.03/s, PixVerse Platform API $0.04/s, WaveSpeedAI $0.09/s.
	T2V I2V F+L	1080p	8 s	$0.03/sfal.ai	2	1	—	—
PixVerse's V3.5 model with transitions between two frames and a fast mode. API: fal.ai $0.03/s, Replicate. Apps: Pollo AI Lite $15/mo.
	Lips	4K	—	$0.133/sfal.ai	2	1	Apr 2026	—
sync.'s most capable lip-sync model, up to 4K at 60 fps. API: fal.ai $0.133/s, WaveSpeedAI $0.135/s. Apps: Higgsfield $0.441/s (Plus $49/mo).
	I2V F+L Ref Extend	1080p	—	$0.051/sVidu API	2	1	—	—
Q2 quality tier for image-to-video, first/last frame, reference and extension. API: Vidu API $0.051/s, fal.ai $0.075/s. Apps: Pollo AI Lite $15/mo.
	I2V F+L Extend	1080p	—	$0.013/sVidu API	2	1	—	—
Faster Q2 for image-to-video and first/last frame. API: Vidu API $0.013/s, fal.ai $0.05/s. Apps: Pollo AI Lite $15/mo.
	Ref Audio	1080p	16 s	$0.07/sfal.ai	2	1	—	—
Vidu's top reference-to-video model for multi-subject consistency. API: fal.ai $0.07/s, Vidu API $0.112/s. Apps: Pollo AI Lite $15/mo.
	T2V I2V	1080p	5 s	$0.02/sAlibaba Cloud Model Studio	2	1	Jul 2025	—
Hosted Wan 2.2 for silent 5 s text- and image-to-video at 480P or 1080P. API: Alibaba Cloud Model Studio $0.02/s, AI/ML API $0.10 per undefined s clip. Apps: Pollo AI Lite $15/mo.
	T2V I2V	480p	6 s	—	2	0	Aug 2024	—
Apache-2.0 2024 open video models from Zhipu. API: ModelsLab $0.05 per undefined s clip, Replicate $0.50 per undefined s clip.
	T2V I2V Lips Audio	1080p	30 s	$0.02/sWaveSpeedAI	2	0	Mar 2026	—
Open 15B human-centric model that generates speech and video together with tight lip sync. API: WaveSpeedAI $0.02/s, fal.ai $0.05/s.
	V2V Edit	720p	—	$0.02/sfal.ai	2	0	—	—
Previous Lucy video editing model, batch or real time at 720p. API: fal.ai $0.02/s, Decart API $0.04/s.
	V2V	720p	—	$0.01/sDecart API	2	0	—	—
Video restyling and real-time style transfer at 720p. API: Decart API $0.01/s, fal.ai $0.01/s.
	I2V F+L	720p	—	$0.03/sPiAPI	2	0	Apr 2025	—
Open next-frame-prediction model that animates an image into long clips (up to about a minute) on modest GPUs. API: PiAPI $0.03/s, fal.ai $0.033/s.
	I2V Lips Audio	1080p	600 s	$0.025/sHedra API	1	1	—	—
Hedra's talking-character model: animates a face image to speech, clips up to 10 minutes. API: Hedra API $0.025/s. Apps: Hedra Studio Basic $20/mo.
	V2V Motion Swap Edit	1080p	30 s	$0.318/seach::labs (Eachlabs)	2	0	Aug 2026	—
Swaps a character, outfit, product or object in a 4-30 s clip, or transfers its motion to a new character, at up to 1080p. API: each::labs (Eachlabs) $0.318/s, Higgsfield API $0.318/s.
	V2V Motion Edit	1080p	—	$0.05/seach::labs (Eachlabs)	2	0	Oct 2026	—
Cheaper, automatic version of Genjutsu object and character swap. API: each::labs (Eachlabs) $0.05/s, Higgsfield API $0.05/s.
	Lips	—	15 s	$0.172/sSegmind	1	1	—	—
Higgsfield's speech-to-video model: a photo plus audio becomes a lip-synced talking clip of 5, 10 or 15 s. API: Segmind $0.172/s. Apps: Higgsfield $0.137/s (Plus $49/mo).
	T2V I2V	720p	—	$0.02/sWaveSpeedAI	2	0	Nov 2025	—
Lightweight open video model from Tencent for text- and image-to-video on consumer GPUs. API: WaveSpeedAI $0.02/s, fal.ai $0.075/s.
	T2V I2V	—	10 s	$0.016/sAI/ML API	2	0	Sep 2025	—
Sber's open 2B text-to-video model; 5 or 10 s clips that run on a 12 GB GPU. API: AI/ML API $0.016/s, fal.ai $0.016/s.
	T2V I2V Camera	1024p	10 s	$0.04/seach::labs (Eachlabs)	2	0	Nov 2025	—
Sber's open text- and image-to-video Pro model, ranked top open text-to-video model on LMArena in December 2025. API: each::labs (Eachlabs) $0.04/s, fal.ai $0.04/s.
	T2V V2V	—	—	$0.025/sAI/ML API	2	0	Oct 2025	—
Krea's open real-time video model based on Wan 14B, for fast text-to-video and live video restyling. API: AI/ML API $0.025/s, fal.ai $0.025/s.
	T2V I2V V2V Extend Edit	4K	10 s	—	0	2	—	—
Previous Ray3 release still offered in Luma's app (and free daily in Adobe Firefly), with HDR and extend. Apps: Luma Dream Machine $0.03/s (Plus $30/mo), Adobe Firefly $0.125/s (Firefly Standard $9.99/mo).
	T2V I2V Extend	—	—	$0.08/sfal.ai	2	0	Apr 2025	—
Apache-2.0 autoregressive video model that generates chunk by chunk for long, extendable videos. API: fal.ai $0.08/s, WaveSpeedAI $0.32 per undefined s clip.
	I2V F+L Extend Loop	720p	21 s	—	0	2	Jun 2025	—
Animates a Midjourney or uploaded image into a 5 second clip, extendable 4 seconds at a time to 21 seconds, with end frame and loop. Apps: Midjourney $0.02/s (Basic $10/mo), Pollo AI Lite $15/mo.
	I2V	720p	5 s	$0.005/sfal.ai	2	0	—	—
TBC's budget image-to-video model, billed as "neuron-derived"; 3 or 5 s clips at 480p or 720p. API: fal.ai $0.005/s, OpenRouter $0.005/s.
	I2V Lips Audio	4K	180 s	$0.02/sPruna API	2	0	—	—
Talking video from one portrait plus audio or a script, 1 second to 3 minutes with synced lips and expressions. API: Pruna API $0.02/s, Replicate $0.02/s.
	T2V I2V F+L Ref Audio Upscale	4K	15 s	$0.01/sPika API	1	1	Oct 2026	—
Pika's post-trained MiniMax H3 (open-weight LoRA adapters): 3 to 15 second clips with synchronized audio in six sampling steps, 480p/768p native with upscaling to 4K. API: Pika API $0.01/s. Apps: Pika app $0.022/s (Starter $10/mo).
	T2V I2V Ref Audio	—	—	$0.11/sWaveSpeedAI	2	0	—	—
Skywork's closed audio-video model, sold through the SkyReels app and API partners. API: WaveSpeedAI $0.11/s, ModelsLab $0.13/s.
	I2V Lips	720p	—	$0.10/sfal.ai	1	1	—	—
Faster-rendering mode of Fabric 1.0 talking video. API: fal.ai $0.10/s. Apps: Freepik (Magnific) Premium €15/mo.
	Motion Swap	—	—	$0.01/sViggle API	1	1	—	—
Viggle's in-house model that puts any character image into a motion video, the engine behind its viral Mix/Remix memes. API: Viggle API $0.01/s. Apps: Viggle app Pro $7.99/mo.
	T2V I2V F+L Ref Edit	720p	5 s	$0.10/sAlibaba Cloud Model Studio	2	0	Feb 2025	—
Hosted Wan 2.1 at 720P, including VACE editing and first/last-frame endpoints. API: Alibaba Cloud Model Studio $0.10/s, AI/ML API $0.50 per undefined s clip.
	I2V F+L	1080p	—	$0.015/sAlibaba Cloud Model Studio	1	1	Jul 2025	—
Cheapest hosted Wan for image-to-video and first/last-frame clips. API: Alibaba Cloud Model Studio $0.015/s. Apps: Pollo AI Lite $15/mo.
	T2V I2V	720p	—	$0.03/sfal.ai	2	0	Jul 2025	—
Compact Apache-2.0 model for 720p text/image-to-video on a single consumer GPU. API: fal.ai $0.03/s, Replicate $0.025 per undefined s clip.
	T2V I2V Camera	1080p	5 s	—	0	1	Feb 2025	—
Adobe's commercially safe video model in the Firefly app; Firefly plans also run partner models (Google, OpenAI, Luma, FLUX and more). Apps: Adobe Firefly $0.10/s (Firefly Standard $9.99/mo).
	T2V	720p	—	$0.035/sDeepInfra	1	0	May 2026	—
Smaller, cheaper sibling of Cosmos 3 Super for world-model video generation. API: DeepInfra $0.035/s.
	T2V I2V	—	—	—	1	0	—	—
NVIDIA's 2B open world model that predicts future video from text or an image, built for robotics and driving simulation. API: WaveSpeedAI $0.30 per undefined s clip.
	Lips	720p	—	$0.07/sfal.ai	1	0	—	—
Studio-quality talking or singing avatar video from one image and an audio track. API: fal.ai $0.07/s.
	V2V Edit	720p	—	$0.04/sDecart API	1	0	—	—
Previous-generation Lucy video editing model. API: Decart API $0.04/s.
	V2V Edit	720p	—	$0.04/sDecart API	1	0	—	—
Decart's latest video editing model, as a batch model or in real time at 720p. API: Decart API $0.04/s.
	V2V Edit	720p	—	$0.04/sDecart API	1	0	—	—
Previous virtual try-on model. API: Decart API $0.04/s.
	V2V Edit	720p	—	$0.04/sDecart API	1	0	—	—
Virtual try-on for video: dresses the person in a clip in new clothes. API: Decart API $0.04/s.
	T2V I2V Ref V2V Lips	—	—	—	0	1	—	—
DomoAI's in-house model for anime-style video-to-video restyles, character video and talking avatars. Apps: DomoAI app Basic $13/mo.
	T2V I2V F+L	—	—	—	0	1	—	—
ByteDance's keyframe-guided one-shot video model built on Seedance 1.0, sold in CapCut. Apps: CapCut Pro.
	Ref V2V	—	—	$0.01/sfal.ai	1	0	Oct 2026	—
Larger, higher-quality EdenLive for real-time generative rendering from a visual reference. API: fal.ai $0.01/s.
	Ref V2V	—	—	$0.0035/sfal.ai	1	0	Oct 2026	—
VAST's open real-time generative renderer that restyles a live video stream from any visual reference. API: fal.ai $0.0035/s.
	T2V	480p	5 s	$0.0025/sDeepInfra	1	0	Jun 2026	—
Distilled, FP8-quantized Wan 1.3B that makes a 5 s 480p clip in seconds on a consumer GPU. API: DeepInfra $0.0025/s.
	T2V I2V	720p	5 s	$0.00225/sDeepInfra	1	0	Aug 2025	—
Sparse-distilled Wan 2.2 TI2V 5B for very fast, very cheap text- and image-to-video. API: DeepInfra $0.00225/s.
	T2V	480p	—	—	1	0	Oct 2024	—
Genmo's open-weight (Apache 2.0) text-to-video model, released as a preview. API: Replicate $0.42 per undefined s clip.
	T2V Ref V2V Audio Camera	1080p	30 s	—	1	0	—	—
Higgsfield's multi-shot film tool: up to 50 image, video and audio references, native audio and directed camera moves, clips up to 30 s. API: Higgsfield API.
	Lips	—	—	$0.03/sWaveSpeedAI	1	0	May 2025	—
Open audio-driven talking and singing avatar model with emotion control and multi-character dialogue. API: WaveSpeedAI $0.03/s.
	I2V	720p	—	$0.08/sfal.ai	1	0	Mar 2025	—
Open image-to-video extension of HunyuanVideo with LoRA effect training. API: fal.ai $0.08/s.
	T2V I2V Audio	480p	5 s	$0.032/sfal.ai	1	0	Sep 2026	—
Fast open Kandinsky 6 model for 5 s clips with sound; optional upscaling to 1080p or 4x. API: fal.ai $0.032/s.
	T2V I2V Audio	480p	5 s	$0.27/sfal.ai	1	0	Sep 2026	—
Kandinsky Lab's open flagship for cinematic 5 s clips with sound; optional upscaling to 1080p or 4x. API: fal.ai $0.27/s.
	Swap Edit	1080p	—	$0.084/sKling AI API	1	0	Jul 2025	—
Adds, swaps or removes elements (people, objects) in an existing video. API: Kling AI API $0.084/s.
	I2V	480p	—	—	0	1	—	—
Quicker 480p mode of Motion 2.0, Leonardo's default for image-to-video. Apps: Leonardo.Ai Essential $12/mo.
	T2V I2V Extend	720p	—	$0.025/sfal.ai	1	0	Oct 2025	—
Meituan's open model built for long videos: it continues clips to reach minutes-long output. API: fal.ai $0.025/s.
	Lips	720p	64 s	$0.04/sWaveSpeedAI	1	0	May 2026	—
Open audio-driven avatar model: one photo plus audio becomes a talking or singing clip up to 64 s. API: WaveSpeedAI $0.04/s.
	Ref V2V Edit	720p	—	$0.04/sReplicate	1	0	—	—
Decart's text- and reference-image-guided video editor at 720p with no length limit. API: Replicate $0.04/s.
	T2V I2V	720p	10 s	$0.06/sWaveSpeedAI	1	0	—	—
Luma's 2024 Dream Machine model; Luma's own API now lists only Ray2 and newer, so it is sold through resellers. API: WaveSpeedAI $0.06/s.
	T2V I2V Ref Motion Camera	—	—	$0.30/sfal.ai	1	0	Jul 2025	—
Video model trained only on licensed footage, built for filmmakers, with pose transfer, motion transfer, camera and trajectory control. API: fal.ai $0.30/s.
	Lips	—	—	$0.03/sWaveSpeedAI	1	0	Jun 2025	—
Open audio-driven model for single- or multi-person talking and singing video from one image. API: WaveSpeedAI $0.03/s.
	Motion Swap	1080p	—	$0.03/sReplicate	1	0	—	—
Animates a reference image with the motion and audio of a source video. API: Replicate $0.03/s.
	T2V I2V	720p	10 s	$0.04/sWaveSpeedAI	1	0	—	—
Faster, cheaper variant of Pika 2.0. API: WaveSpeedAI $0.04/s.
	T2V I2V	720p	10 s	$0.04/sWaveSpeedAI	1	0	—	—
Pika's 2.1 model for sharper 720p text- and image-to-video. API: WaveSpeedAI $0.04/s.
	V2V Edit	—	—	—	1	0	—	—
Adds a described subject into an existing video. API: Pika API.
	I2V V2V	—	—	—	1	0	—	—
Applies a named preset effect to an image or video. API: Pika API.
	V2V Swap Edit	—	—	—	1	0	—	—
Replaces a region, object or person in an existing video. API: Pika API.
	Lips	1080p	—	$0.022/sPixVerse Platform API	1	0	—	—
Talking avatar from a portrait plus uploaded audio or TTS; separate lip-sync endpoint for existing videos. API: PixVerse Platform API $0.022/s.
	T2V I2V F+L	—	10 s	—	0	1	—	—
Pollo AI's own 2025 video model, 5 or 10 s clips with an optional end frame. Apps: Pollo AI Lite $15/mo.
	T2V I2V F+L	1080p	10 s	—	0	1	—	—
Pollo AI's own model, 5 or 10 s clips up to 1080p with an optional end frame. Apps: Pollo AI Lite $15/mo.
	T2V I2V Ref Audio	1080p	10 s	—	0	1	Nov 2025	—
Pollo AI's model with synced sound effects and music, 5 or 10 s clips up to 1080p. Apps: Pollo AI $0.075/s (Lite $15/mo).
	T2V I2V Audio	1080p	15 s	—	0	1	—	—
Pollo AI's multi-shot model with audio, 4-15 s clips at 720p or 1080p. Apps: Pollo AI $0.037/s (Lite $15/mo).
	T2V I2V Ref	4K	15 s	—	0	1	—	—
Pollo AI's newest in-house model in its app, up to 15 s and 4K. Apps: Pollo AI Lite $15/mo.
	T2V Ref	1080p	30 s	—	0	1	—	—
Pollo AI's in-app model for longer cinematic clips up to 30 s. Apps: Pollo AI Lite $15/mo.
	Lips	—	—	$0.0033/sRunway API	1	0	—	—
Talking-character videos from a text script or audio file, plus real-time conversational avatars. API: Runway API $0.0033/s.
	Motion Swap	704p	10 s	$0.20/sfal.ai	1	0	Jun 2026	—
Z.ai's open character-animation model that drives a reference character with a source video, without skeleton maps. API: fal.ai $0.20/s.
	Ref Lips	—	—	$0.03/sWaveSpeedAI	1	0	Jan 2026	—
Open multimodal SkyReels with reference-to-video (R2V 14B) and audio-to-video talking avatars (A2V 19B). API: WaveSpeedAI $0.03/s.
	I2V	—	—	—	1	0	Nov 2023	—
Stability's open-weight image-to-video models (img2vid, img2vid-xt, xt-1.1) for local use. API: ModelsLab $0.05 per undefined s clip.
	Motion	720p	—	$0.047/sVidu API	1	0	—	—
Vidu's motion-transfer endpoint (V1, V2 motion_control_2 and 2.5). API: Vidu API $0.047/s.
	I2V F+L	1080p	—	$0.015/sVidu API	1	0	—	—
Cheapest Q2 image-to-video tier. API: Vidu API $0.015/s.
	I2V Audio	1080p	16 s	$0.056/sVidu API	1	0	—	—
Quicker Q3 Pro for image-to-video. API: Vidu API $0.056/s.
	Lips Edit	—	—	$0.0047/sVidu API	1	0	—	—
Vidu's digital-human model for real-time and offline talking avatars, plus an editing edition. API: Vidu API $0.0047/s.
	Motion Swap	—	—	$0.06/sViggle API	1	0	—	—
Character-replacement video that keeps the driving performance, camera and timing (open-source base fine-tuned by Viggle). API: Viggle API $0.06/s.
	T2V I2V	720p	5 s	$0.036/sAlibaba Cloud Model Studio	1	0	Feb 2025	—
Fast hosted Wan 2.1 at a flat low rate. API: Alibaba Cloud Model Studio $0.036/s.
	T2V I2V	1080p	10 s	—	1	0	Aug 2025	—
ByteDance research model for text- and image-to-video; renders at 720p and upscales to 1080p. API: WaveSpeedAI $0.30 per undefined s clip.
	T2V I2V	—	—	—	0	0	Nov 2024	—
Zhipu's open 5B text- and image-to-video model.
	V2V Edit	—	—	—	0	0	Sep 2025	—
Open-weight version of Decart's Lucy text-guided video editing model (1.1 released Nov 2025).
	Ref Swap Lips	—	—	—	0	0	May 2025	—
Open subject-consistent video generation from image, audio and video conditions, including object replacement.
	Lips	—	—	—	0	0	Sep 2026	—
Lighter open Kandinsky speech-to-video model for talking avatars from a reference frame and speech.
	Lips	—	—	—	0	0	Sep 2026	—
Open speech-to-video model that animates the whole figure in a reference frame to match a speech track.
	T2V I2V	720p	—	—	0	0	Apr 2026	—
Open 2B text- and image-to-video model that makes 121-frame 720p clips from one checkpoint.
	I2V Audio	—	—	—	0	0	Aug 2025	—
Baidu's enterprise image-to-video model family (Turbo/Pro/Lite) with audio; no public self-serve price found.
	T2V I2V Lips Audio	—	—	—	0	0	May 2026	—
Baidu's open 6.3B model that generates video and synchronized audio, including multi-speaker speech, from one prompt.
	T2V I2V Extend	720p	—	—	0	0	Apr 2025	—
Open 'infinite-length' film model using diffusion forcing for long, extendable shots.
	T2V	—	—	—	0	0	Feb 2025	—
StepFun's MIT-licensed 30B open text-to-video model.
	I2V	—	—	—	0	0	Mar 2025	—
StepFun's open image-to-video companion to Step-Video-T2V.
	Motion Swap	—	—	—	0	0	Jul 2026	—
Second open Wan Animate checkpoint: drives a character image with a reference video, or swaps the character into it.
	I2V Motion	—	—	—	0	0	Jul 2026	—
Apache-2.0 image-to-video model from the Wan team for dance and motion videos.
	T2V I2V Ref V2V Audio Edit	4K	10 s	$0.079/sKie.ai	8	4	May 2026	—
First Gemini Omni model (announced at I/O, 19 May 2026); the Vertex preview gemini-omni-flash-preview has a 6 Nov 2026 deprecation date, replaced by Omni 1.1 Flash. API: Kie.ai $0.079/s, Runway API $0.10/s, Segmind $0.10/s, Gemini Enterprise Agent Platform (Vertex AI) $0.101/s, ModelsLab $0.12/s, fal.ai $0.125/s, WaveSpeedAI $0.13/s, each::labs (Eachlabs). Apps: Krea $0.261/s (Basic $9/mo), Adobe Firefly Firefly Standard $9.99/mo, Figma Weave (Weavy) Starter $24/mo, Pollo AI Lite $15/mo.
	T2V I2V F+L Audio Extend	—	—	$0.011/sfal.ai	10	2	Dec 2025	—
First Seedance with native audio; ModelArk retires it on 2026-11-11. BytePlus lists seedance-1-5-pro-251215 under 'Retiring models', retired on November 11, 2026. API: fal.ai $0.011/s, OpenRouter $0.011/s, WaveSpeedAI $0.012/s, Replicate $0.013/s, BytePlus ModelArk $0.025/s, ModelsLab $0.03/s, AI/ML API, DeepInfra, each::labs (Eachlabs), Segmind. Apps: Higgsfield $0.015/s (Starter $15/mo), Freepik (Magnific) Premium €15/mo.
	T2V I2V Audio	1080p	—	$0.12/sPiAPI	6	5	Jul 2025	—
First Veo with native audio; veo-3.0-generate-001 was retired on Vertex AI on 30 June 2026 and Google points users to veo-3.1-generate-001. API: PiAPI $0.12/s, each::labs (Eachlabs) $0.20/s, Replicate $0.20/s, Segmind $0.20/s, WaveSpeedAI $0.20/s, DeepInfra $0.40/s. Apps: Higgsfield $0.355/s (Plus $49/mo), Canva Pro, Freepik (Magnific) Premium €15/mo, Leonardo.Ai Essential $12/mo, Runway (app) Standard $15/mo.
	T2V I2V Audio	—	—	$0.08/sPiAPI	7	3	Sep 2025	—
OpenAI closed the Sora app and website on 26 April 2026 and removed sora-2 from the API on 24 September 2026; its last API price was $0.10/s at 720p. API: PiAPI $0.08/s, each::labs (Eachlabs) $0.10/s, Replicate $0.10/s, WaveSpeedAI $0.10/s, ModelsLab $0.125/s, AI/ML API $0.13/s, Segmind $0.143/s. Apps: CapCut Pro, Freepik (Magnific) Premium €15/mo, Invideo Pro $30/mo.
	T2V I2V Audio	1080p	—	$0.06/sPiAPI	6	4	Jul 2025	—
First Veo with native audio; veo-3.0-fast-generate-001 was retired on Vertex AI on 30 June 2026 and Google points users to veo-3.1-fast-generate-001. API: PiAPI $0.06/s, each::labs (Eachlabs) $0.10/s, Replicate $0.10/s, Segmind $0.10/s, WaveSpeedAI $0.10/s, DeepInfra $0.15/s. Apps: Higgsfield $0.135/s (Plus $49/mo), Poe $0.25/s, Freepik (Magnific) Premium €15/mo, Leonardo.Ai Essential $12/mo.
	I2V F+L	1080p	10 s	$0.025/sKie.ai	5	4	Jun 2025	—
Image-to-video Kling from mid-2025 in Standard (720p) and Pro (1080p) modes; first/last frame in Pro. Not listed on Kling's current API pricing or capability map (checked 2026-11-05); treat as legacy. API: Kie.ai $0.025/s, WaveSpeedAI $0.05/s, ModelsLab $0.056/s, Segmind $0.075/s, Magnific API (formerly Freepik API). Apps: Poe $0.051/s, Higgsfield $0.075/s (Starter $15/mo), Freepik (Magnific) Premium €15/mo, Kaiber Creator $29/mo.
	T2V I2V	1080p	10 s	$0.16/sKie.ai	6	3	Jun 2025	—
Top-quality 2025 Kling tier for text- and image-to-video. Not listed on Kling's current API pricing or capability map (checked 2026-11-05); treat as legacy. API: Kie.ai $0.16/s, PiAPI $0.192/s, WaveSpeedAI $0.26/s, Replicate $0.28/s, ModelsLab $0.336/s, Segmind $0.42/s. Apps: Higgsfield $0.375/s (Starter $15/mo), Freepik (Magnific) Premium €15/mo, Kaiber Creator $29/mo.
	T2V I2V Audio	—	—	$0.24/sPiAPI	6	1	Sep 2025	—
Higher-quality Sora 2 tier, removed from the API on 24 September 2026; last price $0.30/s at 720p, $0.50 at 1024p, $0.70 at 1080p. API: PiAPI $0.24/s, each::labs (Eachlabs) $0.30/s, Replicate $0.30/s, WaveSpeedAI $0.30/s, ModelsLab $0.36/s, Segmind $0.429/s. Apps: Freepik (Magnific) Premium €15/mo.
	T2V I2V F+L Ref	—	—	$0.01/sKie.ai	4	1	Jun 2025	—
2025 budget Seedance tier, no longer on BytePlus ModelArk. Not in ModelArk's model list or pricing (checked 2026-11-05). API: Kie.ai $0.01/s, WaveSpeedAI $0.016/s, Replicate $0.018/s, Together AI $0.029/s. Apps: Freepik (Magnific) Premium €15/mo.
	T2V I2V F+L	—	—	$0.03/sfal.ai	3	1	—	—
2025 PixVerse release, no longer on the PixVerse API capability map. Not listed on PixVerse's API capability map or pricing (checked 2026-11-05). API: fal.ai $0.03/s, WaveSpeedAI $0.05/s, Replicate $0.06/s. Apps: Pollo AI Lite $15/mo.
	Ref V2V Edit	—	—	$0.15/seach::labs (Eachlabs)	4	0	Aug 2025	—
First Aleph video editing model; removed from the Runway API on 30 July 2026 in favour of Aleph 2.0. API: each::labs (Eachlabs) $0.15/s, ModelsLab $0.18/s, Replicate $0.18/s, Segmind $0.356/s.
	T2V I2V F+L Extend Camera	720p	—	$0.50/seach::labs (Eachlabs)	3	1	May 2025	—
Google's silent 720p Veo generation, retired on Vertex AI on 30 June 2026 (veo-2.0-generate-001); replacement veo-3.1-generate-001. API: each::labs (Eachlabs) $0.50/s, Replicate $0.50/s, Segmind $0.50/s. Apps: Freepik (Magnific) Premium €15/mo.
	T2V I2V F+L Ref Extend	1080p	10 s	$0.045/sWaveSpeedAI	3	0	—	—
Early Kling with Standard/Pro modes, end-frame control and 4-5 s video extension. Not listed on Kling's current API pricing or capability map (checked 2026-11-05); treat as legacy. API: WaveSpeedAI $0.045/s, PiAPI $0.052/s, ModelsLab $0.056/s.
	T2V I2V Camera	—	—	$0.056/sTogether AI	3	0	Feb 2025	—
Early-2025 MiniMax models built for precise camera-movement instructions. Not on MiniMax's current pricing page (checked 2026-11-05). API: Together AI $0.056/s, fal.ai $0.083/s, Replicate $0.50 per undefined s clip.
	T2V I2V	—	10 s	$0.26/sWaveSpeedAI	1	1	May 2025	—
Kling 2.0's single high-quality tier (no Std/Pro modes). Not listed on Kling's current API pricing or capability map (checked 2026-11-05); treat as legacy. API: WaveSpeedAI $0.26/s. Apps: Poe $0.18/s.
	I2V	—	—	—	0	1	—	—
Removed from the Runway API on 30 July 2026; Runway points users to Gen-4.5 or Gen-4 Turbo. Apps: Krea $0.10/s (Basic $9/mo).
	T2V I2V	720p	6 s	—	0	0	—	—
First Nova Reel; reached end of life on Bedrock on 30 Sep 2026.
	T2V I2V	720p	120 s	—	0	0	—	—
Amazon's Bedrock video model (6 second shots up to 2 minutes, 1280x720); reached end of life on Bedrock on 30 Sep 2026.
	T2V I2V	1080p	20 s	—	0	0	—	—
OpenAI's first public video model. OpenAI has shut Sora down, and Azure's retirement schedule no longer lists it.

Questions

What is the cheapest AI video generator per second?

Among the 20 models with the highest arena scores, HeyGen Video 1 is the cheapest: $0.01/s on each::labs (Eachlabs), so a 5-second clip costs $0.05. Next come P-Video 2 Pro at $0.01/s on each::labs (Eachlabs) and Grok Imagine Video 1.5 at $0.012/s on Kie.ai. The cheapest model overall is FastWan 2.2 5B at $0.00225/s on DeepInfra, $0.011 per 5-second clip. Prices are the lowest listed rate: silent video at the lowest resolution. Sound and higher resolutions cost more.

What is the best AI video model right now?

Wan 3.0 from Alibaba (Wan, HappyHorse) has the highest arena score in the directory (1172, in preview), ahead of Utopai X (1169) and MiniMax H3 (1160). Arena scores come from blind side-by-side votes on Artificial Analysis. Sort the table by Arena to see the full ranking with prices.

Which AI video models do image-to-video?

145 of the 212 live models animate a still image, including Wan 3.0, Utopai X, MiniMax H3, FLUX 3 Max, Dreamina Seedance 2.5, HeyGen Video 1, MiniMax H3 Max, FLUX 3 Video and 137 more. Pick "Image to video" in the filters to see their prices and where to run them.

Which AI video models generate native audio?

74 live models generate dialogue, effects or music with the video, including Wan 3.0, Utopai X, MiniMax H3, FLUX 3 Max, Dreamina Seedance 2.5, HeyGen Video 1, MiniMax H3 Max, FLUX 3 Video and 66 more. Pick "Native audio" in the filters to compare them.

Where can I use Dreamina Seedance 2.0 by API?

Dreamina Seedance 2.0 runs on 16 APIs: OpenRouter ($0.067/s), BytePlus ModelArk ($0.07/s), AI/ML API ($0.071/s), Segmind ($0.074/s), Replicate ($0.08/s), Kie.ai ($0.095/s), ModelsLab ($0.10/s), PiAPI ($0.10/s), WaveSpeedAI ($0.12/s), Runway API ($0.14/s), Higgsfield API ($0.141/s), fal.ai ($0.173/s), DeepInfra, each::labs (Eachlabs), Magnific API (formerly Freepik API) and Together AI. OpenRouter is the cheapest.

Where can I use MiniMax H3 by API?

MiniMax H3 runs on 15 APIs: Segmind ($0.024/s), PiAPI ($0.03/s), Kie.ai ($0.04/s), fal.ai ($0.05/s), MiniMax API ($0.08/s), Novita AI ($0.08/s), Replicate ($0.08/s), each::labs (Eachlabs) ($0.09/s), ModelsLab ($0.096/s), Runway API ($0.10/s), WaveSpeedAI ($0.10/s), AI/ML API ($0.13/s), Higgsfield API ($0.13/s), OpenRouter ($0.13/s) and Together AI ($0.139/s). Segmind is the cheapest.

Where can I use Dreamina Seedance 2.5 by API?

Dreamina Seedance 2.5 runs on 15 APIs: OpenRouter ($0.103/s), Replicate ($0.103/s), BytePlus ModelArk ($0.103/s), Segmind ($0.109/s), each::labs (Eachlabs) ($0.119/s), ModelsLab ($0.124/s), Kie.ai ($0.14/s), PiAPI ($0.15/s), WaveSpeedAI ($0.18/s), Runway API ($0.20/s), Higgsfield API ($0.206/s), fal.ai ($0.221/s), AI/ML API, Magnific API (formerly Freepik API) and Together AI. OpenRouter is the cheapest.

Which AI video models have open weights?

58 live models ship open weights you can run on your own GPU, including MiniMax H3, HyperFlow H3, MAGI-2 Preview, LTX-2.5 Fast, LTX-2.5 Pro, LTX-2.3 Pro, LTX-2.3 Fast, Pika H3 Flex and 50 more.

How current are these prices?

Every price links to its source and shows the day we checked it. The whole table was last checked on 5 November 2026.