35+ AI video models on the same prompt, compared by cost and render time

This page sends one prompt to 35+ text-to-video models at a matching clip length and shows the clips side by side, with the token cost, clip length and render time for each. The models are the ones in the Creator Studio: Seedance, Kling, Sora, Veo, Wan, LTX, Hailuo, PixVerse, Grok Imagine and more.

The demo prompts are chosen to show motion, not just a pretty frame: a drone move over a landscape, a dog running through surf, and a rainy-street tracking shot. Hover any card to play the preview. Switch prompts, filter by provider, sort by cost or speed, and search by name. Nothing on the demo tab spends your tokens.

To test your own idea, tick the models you want and run your prompt across them. Clip length can be set explicitly or left on Auto, which gives each model its nearest supported length to your target. You see the estimated token total before confirming, every clip is billed at the listed cost, and the results land in your gallery.

Open AI Video Model Comparison on Reality Fabricator

What you can do

35+ text-to-video models compared on the same prompt at a matching clip length
Cost, clip length and render time listed for every model
Demo prompts built around motion: a drone move, a dog in surf, a rainy tracking shot
Hover previews on every card; filter by provider and sort by cost or speed
Run your own prompt across the models you pick, with explicit or automatic clip length
Token estimate before you confirm; every output saved to your gallery

How it works

Choose a demo prompt and hover the cards to play each model's clip.
Filter by provider or sort by cost and speed to find candidates.
Pick the models you want and open the try-it-yourself tab.
Enter your prompt, set a clip length or leave it on Auto, confirm the estimate, and compare.

Frequently asked questions

Which video models are compared?

Every text-to-video model in the Creator Studio, 35+ in total, including Seedance, Kling, Sora, Veo, Wan, LTX, Hailuo, PixVerse and Grok Imagine. You can filter the list by provider and search by name.

Are the clips the same length for every model?

The demo uses a matching clip length across models so the comparison is fair. For your own run, you can set a length or choose Auto, which gives each model its nearest supported length to your target.

Does viewing the demo cost anything?

No. The demo clips are pre-rendered and viewing them spends no tokens. Charges apply only when you run your own prompt, and the page shows the estimated total across your picked models first.

What happens to clips I generate?

Each clip is billed at the listed token cost for that model and saved to your gallery, where you can download it, share it, or open it in the Creator Studio for further work.

Do I need to log in?

Only to run your own prompt. The demo is open to everyone; running a prompt requires an account because outputs are billed in tokens and saved to your gallery.

Related tools

The 42 models compared

Model	Provider	Cost per clip	Best for
Seedance 2.5 T2V (API 4)	atlascloud	51,550 tokens ($1.03) for 5 s	Seedance 2.5 text-to-video via API 4. No content filters. Native 4–30s single pass, up to 50 references, native audio. 480p/720p only (720p is 2.25×, not the 2.0 2× rule).
Seedance 2.5 T2V (API 6)	imagerouter	51,550 tokens ($1.03) for 5 s	Seedance 2.5 text-to-video via API 6. Native 4–30s, 480p/720p only (720p is 2.25×). SFW-routed.
Slight Discount Seedance 2.0 T2V (API 4)	atlascloud	27,700 tokens ($0.55) for 5 s	Fast Seedance 2.0 text-to-video. No content filters. 3x faster than premium. 480p default ~$0.07/sec; 720p ~$0.15/sec. 4-15s. For nudity keep the clip silent — a spoken line dressed her or tripped the audio-copyright block (probed Sep 26 2026).
Premium Seedance 2.0 T2V (API 4)	atlascloud	43,100 tokens ($0.86) for 5 s	Premium Seedance 2.0 text-to-video via API 4. No content filters. Highest cinematic quality. 480p default ~$0.10/sec; 720p ~$0.21/sec. 4-15s. For nudity keep the clip silent — Seedance 2.0 dresses a character who is given a spoken line (probed Sep 26 2026).
Seedance 2.0 Mini T2V (API 1) — Sometimes cheapest, sometimes unusably slow	seegen	15,400 tokens ($0.31) for 5 s	Seedance 2.0 Mini text-to-video via API 1. NSFW capable. 480p/720p native. ~$0.06/sec. For nudity from text keep the clip silent — Seedance 2.0 dresses a character who is given a spoken line (probed Sep 26 2026); Seedance 2.5 does not.
Seedance 2.0 Fast T2V (API 1) — Sometimes cheapest, sometimes unusably slow	seegen	24,650 tokens ($0.49) for 5 s	Seedance 2.0 Fast text-to-video via API 1. NSFW capable. Faster generation, 480p/720p native, multi-reference. ~$0.10/sec. For nudity from text keep the clip silent — Seedance 2.0 dresses a character who is given a spoken line (probed Sep 26 2026); Seedance 2.5 does not.
Seedance 2.0 Pro T2V (API 1) — Sometimes cheapest, sometimes unusably slow	seegen	30,800 tokens ($0.62) for 5 s	Seedance 2.0 Pro text-to-video via API 1. NSFW capable. Native 4K, multi-reference, first+last frame, native audio. ~$0.13/sec. For nudity from text keep the clip silent — Seedance 2.0 dresses a character who is given a spoken line (probed Sep 26 2026); Seedance 2.5 does not.
Seedance 2.0 - AI Video Generator with Audio (API 2)	replicate	69,250 tokens ($1.39) for 5 s	Cinematic AI video generation with native audio, 2K resolution, and camera control. Generates 5-second video clips from text prompts.
Wan 3.0 T2V (API 1) — 2-30s, native audio, 1080p	seegen	32,850 tokens ($0.66) for 5 s	Alibaba Wan 3.0 text-to-video via API 1. 2-30 second clips with synchronized speech/SFX/music, native 480p/720p/1080p. Prompt is auto-enriched. ~$0.13/sec at 480p (720p 2x, 1080p 4x). Alibaba's content filter applies (refuses nudity; the same model is uncensored on API 4 / API 5).
Wan 3.0 T2V - 2-30s (API 5)	venice	27,950 tokens ($0.56) for 5 s	Wan 3.0 (uncensored) text-to-video via API 5: 2/5/10/15/20/25/30 second clips at 480p (cheapest), 720p or 1080p, native audio. Nudity renders (probed Sep 26 2026); light content filters: Venice still refuses some prompts. Private generation (not logged).
Wan 2.7 T2V (with audio) (API 4)	atlascloud	38,500 tokens ($0.77) for 5 s	Official Wan 2.7 — best overall video quality with native audio output. 1080p, multi-shot. Uncensored but not NSFW-specialized. Use detailed 80-120 word prompts.
Wan 2.7 Enhanced T2V (API 5)	venice	54,200 tokens ($1.08) for 5 s	Wan 2.7 Enhanced (uncensored) text-to-video via API 5 — the successor of the retired Wan 2.7 Uncensored id. Light content filters (Venice still refuses some prompts), native audio, 5/10/15 second clips at 720p or 1080p. Private generation (not logged).
FLUX 3 Video T2V (API 6)	imagerouter	65,400 tokens ($1.31) for 5 s	Black Forest Labs FLUX 3 Video text-to-video via API 6 — up to 20s with native multilingual dialogue and sound effects. HD ($0.17/sec); Full HD 1.71×. SFW-routed.
MiniMax H3 / Hailuo 3.0 T2V (API 6)	imagerouter	50,000 tokens ($1.00) for 5 s	MiniMax Hailuo 3.0 text-to-video via API 6 — native 2K 24fps with single-pass stereo audio, multi-shot 5–15s. Always 2K ($0.13/sec). SFW-routed.
MiniMax H3 Max T2V (API 6)	imagerouter	19,250 tokens ($0.39) for 5 s	MiniMax Hailuo 3.0 Max text-to-video via API 6 — the speed-tuned H3 tier with native stereo audio, 5-15s, 480p ($0.08/sec) or 768p-class (1.6x). SFW-routed.
MiniMax H3 Max Turbo T2V (API 6)	imagerouter	9,650 tokens ($0.19) for 5 s	MiniMax Hailuo 3.0 Max Turbo text-to-video via API 6 — H3 Max at half the price, native stereo audio, 5-15s, 480p ($0.04/sec) or 768p-class (1.6x). SFW-routed.
MiniMax H3 Fast T2V (API 6)	imagerouter	17,700 tokens ($0.35) for 5 s	MiniMax Hailuo 3.0 Fast text-to-video via API 6 — the 480p preview tier with native stereo audio, 5-15s ($0.07/sec). SFW-routed.
MiniMax H3 Max T2V (API 4) — 480P/768P, native audio, fast	atlascloud	18,300 tokens ($0.37) for 5 s	MiniMax Hailuo 3.0 Max text-to-video via API 4 — the speed-tuned H3 tier: native stereo audio, 5-15s, 480P (cheapest) / 768P (1.6x), a 5 s clip in about 10 s. ~$0.07/sec at 480P. Same uncensored route as H3 (nudity renders on this family, probed Sep 26 / Oct 22 2026).
MiniMax H3 Fast T2V (API 4) — 480P, native audio, fast	atlascloud	16,850 tokens ($0.34) for 5 s	MiniMax Hailuo 3.0 Fast text-to-video via API 4 — the 480P preview tier with native stereo audio, 5-15s, a 5 s clip in about 12 s. ~$0.07/sec. Same uncensored route as H3 (probed Sep 26 / Oct 22 2026).
MiniMax H3 Developer T2V (API 4) — cheapest native-audio video	atlascloud	5,800 tokens ($0.12) for 5 s	MiniMax Hailuo 3.0 Developer text-to-video via API 4 — the budget H3 tier: native stereo audio, 4-15s, 480P (cheapest) / 768P (1.6x). ~$0.02/sec at 480P, the cheapest native-audio video on the site. Same uncensored route as H3 (nudity rendered, probed Oct 22 2026).
Kling 3.0 - Cinematic T2V + Audio up to 15s (API 2)	replicate	129,250 tokens ($2.58) for 5 s	Latest Kling 3.0: text-to-video up to 15 seconds with native audio (dialogue lip-sync, SFX, ambient). Multi-shot mode (up to 6 scenes). Standard (720p) or Pro (1080p). ~$0.10/sec.
Gemini Omni 1.1 Flash - T2V/I2V + Audio (Google)	gemini	77,000 tokens ($1.54) for 8 s	Google flagship video generation with conversational editing (served by Omni 1.1 Flash, the GA successor of the preview). Text-to-video and image-to-video with audio. Fixed-length ~10s clips (the model decides — duration is not controllable), 720p. Billed as one 10s clip.
PixVerse V6 - Cinematic T2V + Audio up to 15s (API 2)	replicate	46,200 tokens ($0.92) for 5 s	Top-ranked T2V alongside Seedance 2.0. Multi-shot, native audio (dialogue, SFX), precise camera control, native text rendering. 720p/1080p, 5-15s. ~$0.12/sec (720p+audio).
Hailuo 02 - Best Physics T2V (API 2)	replicate	20,820 tokens ($0.42) for 6 s	MiniMax Hailuo 02: excellent physics simulation, complex motion, 3× bigger than predecessor. Text-only — no source image needed. 768p, 6/10s. ~$0.03/sec (std est.).
LTX-2 v2.3 Fast - 20s T2V 1080p-4K (API 5)	venice	33,900 tokens ($0.68) for 6 s	Lightricks LTX-2 v2.3 Fast (the successor of LTX 2.0 Fast): 6-20 second clips (even seconds), 1080p / 1440p / 4K, native audio. Fast generation with good quality.
Grok Imagine Video - T2V (xAI)	grok	19,250 tokens ($0.39) for 5 s	xAI text-to-video generation. 5/10/15s duration, 480p ($0.05/sec) or 720p ($0.07/sec).
Kling 3.0 Turbo T2V (API 6)	imagerouter	43,100 tokens ($0.86) for 5 s	Kling 3.0 Turbo text-to-video via API 6 — fast Kling 3.0 tier, 3–15s, 720p ($0.112/sec) or 1080p (1.25×). SFW-routed (Kling blocks nudity).
LTX-2.5 Pro T2V (API 6)	imagerouter	55,440 tokens ($1.11) for 6 s	Lightricks LTX-2.5 Pro text-to-video via API 6 — the quality LTX tier, 6/8/10s, 720p ($0.12/sec) or 1080p (2.5×). SFW-routed.
Veo 3.1 Lite T2V (API 6)	imagerouter	15,400 tokens ($0.31) for 4 s	Google Veo 3.1 Lite text-to-video via API 6 — the cheapest Veo with audio, 4/6/8s, 720p ($0.05/sec) or 1080p (4×). SFW-routed.
LTX-2.5 Fast T2V (API 6)	imagerouter	41,580 tokens ($0.83) for 6 s	Lightricks LTX-2.5 Fast text-to-video via API 6 — very fast, 6–20s in 2s steps, 720p ($0.09/sec) or 1080p (1.67×). SFW-routed.
Wan 3.0 Prime T2V (API 1) — faster Wan 3.0	seegen	45,150 tokens ($0.90) for 5 s	Alibaba Wan 3.0 Prime text-to-video via API 1: same inputs as Wan 3.0, tuned for faster generation. 2-30s, native audio, native 480p/720p/1080p. ~$0.18/sec at 480p (720p 2x, 1080p 4x). Alibaba's content filter applies (refuses nudity; uncensored on API 4 / API 5).
Van 2.5 T2V Cinematic + Audio (API 4)	atlascloud	26,200 tokens ($0.52) for 5 s	Cinematic uncensored T2V with native audio sync. 720p/1080p, landscape or portrait, 5-10s. Stable motion + prompt-faithful storytelling. $0.068/sec at 720p (~$0.34/5s), 1.5× at 1080p.
Vidu Q3 Turbo T2V Cheapest (API 4)	atlascloud	31,400 tokens ($0.63) for 5 s	Cheapest uncensored text-to-video. 1080p, audio, multiple styles. Good quality at $0.082/sec (~$0.41/5s). Great for bulk NSFW video generation.
Vidu Q3 Pro T2V Quality (API 4)	atlascloud	40,900 tokens ($0.82) for 5 s	Higher quality uncensored text-to-video. 1080p, audio, smooth motion. ~$0.10/sec (~$0.50/5s). Better detail than Turbo variant.
PixVerse C1 - Latest T2V (API 5)	venice	18,350 tokens ($0.37) for 5 s	PixVerse C1: latest generation with improved motion quality and character consistency. 3/5/8/10/15 second clips, 360p-1080p, audio.
LTX-2 v2.3 Full - Premium 10s T2V 1080p-4K (API 5)	venice	44,880 tokens ($0.90) for 6 s	Lightricks LTX-2 v2.3 Full quality: maximum fidelity, 6/8/10 second clips, 1080p / 1440p / 4K, native audio. Slower but highest quality.
LTX-2.5 Fast - 20s T2V 720p-4K (API 5)	venice	49,920 tokens ($1.00) for 6 s	Lightricks LTX-2.5 Fast (Aug 2026): 6-20 second clips (even seconds), 720p / 1080p / 1440p / 4K, native audio. The current-generation LTX at the fast tier.
LTX-2.5 Pro - 10s T2V 720p/1080p (API 5)	venice	66,900 tokens ($1.34) for 6 s	Lightricks LTX-2.5 Pro (Aug 2026): highest LTX quality, 6/8/10 second clips, 720p / 1080p, native audio.
Kling 3.0 Omni - Unified T2V/I2V/Edit + Refs (API 2)	replicate	86,160 tokens ($1.72) for 5 s	Unified model: T2V, I2V, video editing, style transfer. Up to 7 reference images for character consistency. Native audio, 3-15s. 720p by default; 1080p with resolution "1080p". Per second: 12,924 tokens at 720p silent, 17,232 at 720p with sound or 1080p silent, 21,540 at 1080p with sound.
Longcat - 30s Single-Shot T2V (API 5)	venice	21,200 tokens ($0.42) for 5 s	Longest single-shot video available: 5/10/15/20/30 second clips, 720p. Excellent for long scenes, slow motion, and extended narratives.
Longcat Distilled - Fast 30s T2V (API 5)	venice	7,650 tokens ($0.15) for 5 s	Faster Longcat variant: 5/10/15/20/30 second clips at 720p with reduced generation time. Slightly lower quality but much faster and cheaper.
AnimateDiff - Text-to-Animation/GIF generator (API 2)	replicate	20,000 tokens ($0.40) for 5 s	Creates animated GIFs and videos from text prompts - generates motion sequences

The test prompts and what each model made

Camera move over a landscape

Slow cinematic drone shot flying forward over a misty pine forest at sunrise, a river winding through the valley below, golden light breaking through low clouds. Smooth, steady forward motion.

What to look for: Steady camera motion without warping, trees and river that stay consistent as the shot moves, believable light and mist.

42 models made it; the fastest was MiniMax H3 Fast T2V (API 4) — 480P, native audio, fast (13 s).

Seedance 2.5 T2V (API 4): 51,550 tokens ($1.03), made in 146 s, 5.0 s with sound, 1280×720
Seedance 2.5 T2V (API 6): 51,550 tokens ($1.03), made in 516 s, 5.0 s with sound, 854×480
Slight Discount Seedance 2.0 T2V (API 4): 27,700 tokens ($0.55), made in 148 s, 5.1 s with sound, 1280×720
Premium Seedance 2.0 T2V (API 4): 43,100 tokens ($0.86), made in 117 s, 5.1 s with sound, 1280×720
Seedance 2.0 Mini T2V (API 1) — Sometimes cheapest, sometimes unusably slow: 15,400 tokens ($0.31), made in 129 s, 5.1 s with sound, 864×496
Seedance 2.0 Fast T2V (API 1) — Sometimes cheapest, sometimes unusably slow: 24,650 tokens ($0.49), made in 129 s, 5.1 s with sound, 864×496
Seedance 2.0 Pro T2V (API 1) — Sometimes cheapest, sometimes unusably slow: 30,800 tokens ($0.62), made in 129 s, 5.1 s with sound, 864×496
Seedance 2.0 - AI Video Generator with Audio (API 2): 69,250 tokens ($1.39), made in 188 s, 5.1 s with sound, 864×496
Wan 3.0 T2V (API 1) — 2-30s, native audio, 1080p: 32,850 tokens ($0.66), made in 152 s, 5.0 s with sound, 854×480
Wan 3.0 T2V - 2-30s (API 5): 19,050 tokens ($0.38), made in 44 s, 5.0 s with sound, 854×480
Wan 2.7 T2V (with audio) (API 4): 38,500 tokens ($0.77), made in 193 s, 5.0 s with sound, 1280×720
Wan 2.7 Enhanced T2V (API 5): 54,200 tokens ($1.08), made in 72 s, 5.0 s with sound, 1276×720
FLUX 3 Video T2V (API 6): 65,400 tokens ($1.31), made in 80 s, 5.0 s with sound, 1280×704
MiniMax H3 / Hailuo 3.0 T2V (API 6): 50,000 tokens ($1.00), made in 221 s, 5.2 s with sound, 2560×1440
MiniMax H3 Max T2V (API 6): 19,250 tokens ($0.39), made in 20 s, 5.2 s with sound, 832×480
MiniMax H3 Max Turbo T2V (API 6): 9,650 tokens ($0.19), made in 20 s, 5.2 s with sound, 832×480
MiniMax H3 Fast T2V (API 6): 17,700 tokens ($0.35), made in 20 s, 5.2 s with sound, 864×480
MiniMax H3 Max T2V (API 4) — 480P/768P, native audio, fast: 18,300 tokens ($0.37), made in 13 s, 5.2 s with sound, 1344×768
MiniMax H3 Fast T2V (API 4) — 480P, native audio, fast: 16,850 tokens ($0.34), made in 13 s, 5.2 s with sound, 864×480
MiniMax H3 Developer T2V (API 4) — cheapest native-audio video: 5,800 tokens ($0.12), made in 139 s, 5.2 s with sound, 1344×768
Kling 3.0 - Cinematic T2V + Audio up to 15s (API 2): 129,250 tokens ($2.58), made in 132 s, 5.0 s with sound, 1920×1080
Gemini Omni 1.1 Flash - T2V/I2V + Audio (Google): 77,000 tokens ($1.54), made in 35 s, 10.0 s with sound, 1280×720
PixVerse V6 - Cinematic T2V + Audio up to 15s (API 2): 46,200 tokens ($0.92), made in 46 s, 5.0 s with sound, 1280×720
Hailuo 02 - Best Physics T2V (API 2): 20,820 tokens ($0.42), made in 87 s, 5.9 s, 1366×768
LTX-2 v2.3 Fast - 20s T2V 1080p-4K (API 5): 33,900 tokens ($0.68), made in 38 s, 6.1 s with sound, 1920×1080
Grok Imagine Video - T2V (xAI): 19,250 tokens ($0.39), made in 45 s, 5.0 s with sound, 1280×720
Kling 3.0 Turbo T2V (API 6): 43,100 tokens ($0.86), made in 212 s, 5.0 s with sound, 1280×720
LTX-2.5 Pro T2V (API 6): 55,440 tokens ($1.11), made in 39 s, 6.0 s with sound, 1280×720
Veo 3.1 Lite T2V (API 6): 15,400 tokens ($0.31), made in 40 s, 4.0 s with sound, 1280×720
LTX-2.5 Fast T2V (API 6): 41,580 tokens ($0.83), made in 39 s, 6.0 s with sound, 1280×720
Wan 3.0 Prime T2V (API 1) — faster Wan 3.0: 45,150 tokens ($0.90), made in 194 s, 5.0 s with sound, 854×480
Van 2.5 T2V Cinematic + Audio (API 4): 26,200 tokens ($0.52), made in 97 s, 5.0 s with sound, 1264×728
Vidu Q3 Turbo T2V Cheapest (API 4): 31,400 tokens ($0.63), made in 103 s, 5.0 s with sound, 1280×720
Vidu Q3 Pro T2V Quality (API 4): 40,900 tokens ($0.82), made in 445 s, 5.0 s with sound, 1280×720
PixVerse C1 - Latest T2V (API 5): 18,350 tokens ($0.37), made in 32 s, 5.0 s with sound, 640×360
LTX-2 v2.3 Full - Premium 10s T2V 1080p-4K (API 5): 44,880 tokens ($0.90), made in 74 s, 6.1 s with sound, 1920×1080
LTX-2.5 Fast - 20s T2V 720p-4K (API 5): 49,920 tokens ($1.00), made in 27 s, 6.1 s with sound, 1280×720
LTX-2.5 Pro - 10s T2V 720p/1080p (API 5): 66,900 tokens ($1.34), made in 32 s, 6.1 s with sound, 1280×720
Kling 3.0 Omni - Unified T2V/I2V/Edit + Refs (API 2): 107,700 tokens ($2.15), made in 133 s, 5.0 s with sound, 1920×1080
Longcat - 30s Single-Shot T2V (API 5): 21,200 tokens ($0.42), made in 408 s, 8.1 s, 1280×704
Longcat Distilled - Fast 30s T2V (API 5): 7,650 tokens ($0.15), made in 260 s, 8.1 s, 1280×704
AnimateDiff - Text-to-Animation/GIF generator (API 2): 20,000 tokens ($0.40), made in 78 s, 2.1 s, 512×512

Motion and physics

A golden retriever runs along a sandy beach toward the camera, splashing through shallow waves in slow motion. Bright afternoon sun, water droplets sparkling, shallow depth of field.

What to look for: Leg motion and body consistency, how the water and splashes behave, and whether it stays the same dog for the whole clip.

42 models made it; the fastest was MiniMax H3 Fast T2V (API 4) — 480P, native audio, fast (13 s).

Seedance 2.5 T2V (API 4): 51,550 tokens ($1.03), made in 149 s, 5.1 s with sound, 1280×720
Seedance 2.5 T2V (API 6): 51,550 tokens ($1.03), made in 265 s, 5.0 s with sound, 854×480
Slight Discount Seedance 2.0 T2V (API 4): 27,700 tokens ($0.55), made in 96 s, 5.1 s with sound, 1280×720
Premium Seedance 2.0 T2V (API 4): 43,100 tokens ($0.86), made in 114 s, 5.1 s with sound, 1280×720
Seedance 2.0 Mini T2V (API 1) — Sometimes cheapest, sometimes unusably slow: 15,400 tokens ($0.31), made in 118 s, 5.1 s with sound, 864×496
Seedance 2.0 Fast T2V (API 1) — Sometimes cheapest, sometimes unusably slow: 24,650 tokens ($0.49), made in 118 s, 5.1 s with sound, 864×496
Seedance 2.0 Pro T2V (API 1) — Sometimes cheapest, sometimes unusably slow: 30,800 tokens ($0.62), made in 113 s, 5.1 s with sound, 864×496
Seedance 2.0 - AI Video Generator with Audio (API 2): 69,250 tokens ($1.39), made in 112 s, 5.1 s with sound, 864×496
Wan 3.0 T2V (API 1) — 2-30s, native audio, 1080p: 32,850 tokens ($0.66), made in 178 s, 5.0 s with sound, 854×480
Wan 3.0 T2V - 2-30s (API 5): 19,050 tokens ($0.38), made in 42 s, 5.0 s with sound, 854×480
Wan 2.7 T2V (with audio) (API 4): 38,500 tokens ($0.77), made in 87 s, 5.0 s with sound, 1280×720
Wan 2.7 Enhanced T2V (API 5): 54,200 tokens ($1.08), made in 74 s, 5.0 s with sound, 1276×720
FLUX 3 Video T2V (API 6): 65,400 tokens ($1.31), made in 91 s, 5.0 s with sound, 1280×704
MiniMax H3 / Hailuo 3.0 T2V (API 6): 50,000 tokens ($1.00), made in 231 s, 5.2 s with sound, 2560×1440
MiniMax H3 Max T2V (API 6): 19,250 tokens ($0.39), made in 19 s, 5.2 s with sound, 832×480
MiniMax H3 Max Turbo T2V (API 6): 9,650 tokens ($0.19), made in 19 s, 5.2 s with sound, 832×480
MiniMax H3 Fast T2V (API 6): 17,700 tokens ($0.35), made in 29 s, 5.2 s with sound, 864×480
MiniMax H3 Max T2V (API 4) — 480P/768P, native audio, fast: 18,300 tokens ($0.37), made in 13 s, 5.2 s with sound, 1344×768
MiniMax H3 Fast T2V (API 4) — 480P, native audio, fast: 16,850 tokens ($0.34), made in 13 s, 5.2 s with sound, 864×480
MiniMax H3 Developer T2V (API 4) — cheapest native-audio video: 5,800 tokens ($0.12), made in 22 s, 5.2 s with sound, 1344×768
Kling 3.0 - Cinematic T2V + Audio up to 15s (API 2): 129,250 tokens ($2.58), made in 137 s, 5.0 s with sound, 1920×1080
Gemini Omni 1.1 Flash - T2V/I2V + Audio (Google): 77,000 tokens ($1.54), made in 35 s, 10.0 s with sound, 1280×720
PixVerse V6 - Cinematic T2V + Audio up to 15s (API 2): 46,200 tokens ($0.92), made in 46 s, 5.0 s with sound, 1280×720
Hailuo 02 - Best Physics T2V (API 2): 20,820 tokens ($0.42), made in 92 s, 5.9 s, 1366×768
LTX-2 v2.3 Fast - 20s T2V 1080p-4K (API 5): 33,900 tokens ($0.68), made in 38 s, 6.1 s with sound, 1920×1080
Grok Imagine Video - T2V (xAI): 19,250 tokens ($0.39), made in 50 s, 5.0 s with sound, 1280×720
Kling 3.0 Turbo T2V (API 6): 43,100 tokens ($0.86), made in 56 s, 5.0 s with sound, 1280×720
LTX-2.5 Pro T2V (API 6): 55,440 tokens ($1.11), made in 30 s, 6.0 s with sound, 1280×720
Veo 3.1 Lite T2V (API 6): 15,400 tokens ($0.31), made in 40 s, 4.0 s with sound, 1280×720
LTX-2.5 Fast T2V (API 6): 41,580 tokens ($0.83), made in 40 s, 6.0 s with sound, 1280×720
Wan 3.0 Prime T2V (API 1) — faster Wan 3.0: 45,150 tokens ($0.90), made in 231 s, 5.0 s with sound, 854×480
Van 2.5 T2V Cinematic + Audio (API 4): 26,200 tokens ($0.52), made in 92 s, 5.0 s with sound, 1264×728
Vidu Q3 Turbo T2V Cheapest (API 4): 31,400 tokens ($0.63), made in 176 s, 5.0 s with sound, 1280×720
Vidu Q3 Pro T2V Quality (API 4): 40,900 tokens ($0.82), made in 204 s, 5.0 s with sound, 1280×720
PixVerse C1 - Latest T2V (API 5): 18,350 tokens ($0.37), made in 32 s, 5.0 s with sound, 640×360
LTX-2 v2.3 Full - Premium 10s T2V 1080p-4K (API 5): 44,880 tokens ($0.90), made in 90 s, 6.1 s with sound, 1920×1080
LTX-2.5 Fast - 20s T2V 720p-4K (API 5): 49,920 tokens ($1.00), made in 27 s, 6.1 s with sound, 1280×720
LTX-2.5 Pro - 10s T2V 720p/1080p (API 5): 66,900 tokens ($1.34), made in 27 s, 6.1 s with sound, 1280×720
Kling 3.0 Omni - Unified T2V/I2V/Edit + Refs (API 2): 107,700 tokens ($2.15), made in 254 s, 5.0 s with sound, 1920×1080
Longcat - 30s Single-Shot T2V (API 5): 21,200 tokens ($0.42), made in 361 s, 8.1 s, 1280×704
Longcat Distilled - Fast 30s T2V (API 5): 7,650 tokens ($0.15), made in 229 s, 8.1 s, 1280×704
AnimateDiff - Text-to-Animation/GIF generator (API 2): 20,000 tokens ($0.40), made in 139 s, 2.1 s, 512×512

Person and tracking shot

A young woman in a red coat walks through a busy Tokyo street at night in the rain, holding a clear umbrella. Neon signs reflect on the wet ground. The camera tracks alongside her at walking pace.

What to look for: Face and coat consistency while walking, umbrella transparency, the crowd and neon reflections, and whether the tracking shot stays smooth.

42 models made it; the fastest was MiniMax H3 Max T2V (API 4) — 480P/768P, native audio, fast (14 s).

Seedance 2.5 T2V (API 4): 51,550 tokens ($1.03), made in 143 s, 5.1 s with sound, 1280×720
Seedance 2.5 T2V (API 6): 51,550 tokens ($1.03), made in 182 s, 5.0 s with sound, 854×480
Slight Discount Seedance 2.0 T2V (API 4): 27,700 tokens ($0.55), made in 97 s, 5.1 s with sound, 1280×720
Premium Seedance 2.0 T2V (API 4): 43,100 tokens ($0.86), made in 135 s, 5.1 s with sound, 1280×720
Seedance 2.0 Mini T2V (API 1) — Sometimes cheapest, sometimes unusably slow: 15,400 tokens ($0.31), made in 98 s, 5.1 s with sound, 864×496
Seedance 2.0 Fast T2V (API 1) — Sometimes cheapest, sometimes unusably slow: 24,650 tokens ($0.49), made in 118 s, 5.1 s with sound, 864×496
Seedance 2.0 Pro T2V (API 1) — Sometimes cheapest, sometimes unusably slow: 30,800 tokens ($0.62), made in 114 s, 5.1 s with sound, 864×496
Seedance 2.0 - AI Video Generator with Audio (API 2): 69,250 tokens ($1.39), made in 117 s, 5.1 s with sound, 864×496
Wan 3.0 T2V (API 1) — 2-30s, native audio, 1080p: 32,850 tokens ($0.66), made in 141 s, 5.0 s with sound, 854×480
Wan 3.0 T2V - 2-30s (API 5): 19,050 tokens ($0.38), made in 54 s, 5.0 s with sound, 854×480
Wan 2.7 T2V (with audio) (API 4): 38,500 tokens ($0.77), made in 93 s, 5.0 s with sound, 1280×720
Wan 2.7 Enhanced T2V (API 5): 54,200 tokens ($1.08), made in 67 s, 5.0 s with sound, 1276×720
FLUX 3 Video T2V (API 6): 65,400 tokens ($1.31), made in 79 s, 5.0 s with sound, 1280×704
MiniMax H3 / Hailuo 3.0 T2V (API 6): 50,000 tokens ($1.00), made in 213 s, 5.2 s with sound, 2560×1440
MiniMax H3 Max T2V (API 6): 19,250 tokens ($0.39), made in 20 s, 5.2 s with sound, 832×480
MiniMax H3 Max Turbo T2V (API 6): 9,650 tokens ($0.19), made in 20 s, 5.2 s with sound, 832×480
MiniMax H3 Fast T2V (API 6): 17,700 tokens ($0.35), made in 20 s, 5.2 s with sound, 864×480
MiniMax H3 Max T2V (API 4) — 480P/768P, native audio, fast: 18,300 tokens ($0.37), made in 14 s, 5.2 s with sound, 1344×768
MiniMax H3 Fast T2V (API 4) — 480P, native audio, fast: 16,850 tokens ($0.34), made in 18 s, 5.2 s with sound, 864×480
MiniMax H3 Developer T2V (API 4) — cheapest native-audio video: 5,800 tokens ($0.12), made in 22 s, 5.2 s with sound, 1344×768
Kling 3.0 - Cinematic T2V + Audio up to 15s (API 2): 129,250 tokens ($2.58), made in 153 s, 5.0 s with sound, 1920×1080
Gemini Omni 1.1 Flash - T2V/I2V + Audio (Google): 77,000 tokens ($1.54), made in 35 s, 10.0 s with sound, 1280×720
PixVerse V6 - Cinematic T2V + Audio up to 15s (API 2): 46,200 tokens ($0.92), made in 46 s, 5.0 s with sound, 1280×720
Hailuo 02 - Best Physics T2V (API 2): 20,820 tokens ($0.42), made in 112 s, 5.9 s, 1366×768
LTX-2 v2.3 Fast - 20s T2V 1080p-4K (API 5): 33,900 tokens ($0.68), made in 33 s, 6.1 s with sound, 1920×1080
Grok Imagine Video - T2V (xAI): 19,250 tokens ($0.39), made in 50 s, 5.0 s with sound, 1280×720
Kling 3.0 Turbo T2V (API 6): 43,100 tokens ($0.86), made in 153 s, 5.0 s with sound, 1280×720
LTX-2.5 Pro T2V (API 6): 55,440 tokens ($1.11), made in 30 s, 6.0 s with sound, 1280×720
Veo 3.1 Lite T2V (API 6): 15,400 tokens ($0.31), made in 40 s, 4.0 s with sound, 1280×720
LTX-2.5 Fast T2V (API 6): 41,580 tokens ($0.83), made in 29 s, 6.0 s with sound, 1280×720
Wan 3.0 Prime T2V (API 1) — faster Wan 3.0: 45,150 tokens ($0.90), made in 180 s, 5.0 s with sound, 854×480
Van 2.5 T2V Cinematic + Audio (API 4): 26,200 tokens ($0.52), made in 97 s, 5.0 s with sound, 1264×728
Vidu Q3 Turbo T2V Cheapest (API 4): 31,400 tokens ($0.63), made in 76 s, 5.0 s with sound, 1280×720
Vidu Q3 Pro T2V Quality (API 4): 40,900 tokens ($0.82), made in 165 s, 5.0 s with sound, 1280×720
PixVerse C1 - Latest T2V (API 5): 18,350 tokens ($0.37), made in 32 s, 5.0 s with sound, 640×360
LTX-2 v2.3 Full - Premium 10s T2V 1080p-4K (API 5): 44,880 tokens ($0.90), made in 79 s, 6.1 s with sound, 1920×1080
LTX-2.5 Fast - 20s T2V 720p-4K (API 5): 49,920 tokens ($1.00), made in 33 s, 6.1 s with sound, 1280×720
LTX-2.5 Pro - 10s T2V 720p/1080p (API 5): 66,900 tokens ($1.34), made in 33 s, 6.1 s with sound, 1280×720
Kling 3.0 Omni - Unified T2V/I2V/Edit + Refs (API 2): 107,700 tokens ($2.15), made in 133 s, 5.0 s with sound, 1920×1080
Longcat - 30s Single-Shot T2V (API 5): 21,200 tokens ($0.42), made in 430 s, 8.1 s, 1280×704
Longcat Distilled - Fast 30s T2V (API 5): 7,650 tokens ($0.15), made in 164 s, 8.1 s, 1280×704
AnimateDiff - Text-to-Animation/GIF generator (API 2): 20,000 tokens ($0.40), made in 139 s, 2.1 s, 512×512

Back to Reality Fabricator