Skip to main content

Direct comparison

Sora 2 vs. HunyuanVideo 1.5

Compare the exact model versions across shared blind-preference benchmarks, public pricing, release dates, resolution, duration, native audio, access, and model availability.

Overview

Technical specifications and pricing

Technical limits come from model or API documentation. A representative minute price appears only when it can be traced to the exact model. Different resolutions and audio settings may still affect the final charge.

Specifications and pricing for Sora 2 and HunyuanVideo 1.5
AttributeSora 2HunyuanVideo 1.5
ProviderOpenAITencent
ReleasedSeptember 2025November 2025
AccessAPIAPI, Local
WeightsProprietaryOpen weights
ParametersNot published8.3 billion
Maximum resolution720p1080p (VSR)
Maximum duration20 sNot published
Native audioYesNo
Representative API price$6 per minute SourceNot published
LifecycleMarked as legacy. API shutdown announced for September 24, 2026 Status sourceNo separate notice
Technical sourceModel sourceVerified 2026-08-24Model sourceVerified 2026-08-24

Performance

Shared video benchmarks

A row appears only when both exact model versions occur in the same LMArena task and stored dataset revision. Text-to-video and image-to-video are never merged into one score.

Sora 2HunyuanVideo 1.5
LMArena Text to VideoBlind preference rating from the same task and dataset revision
Sora 2Higher value1,340.1
HunyuanVideo 1.51,169.2

Key to the marks:better valuebetter, but not conclusivetie

Shared video benchmarks for Sora 2 and HunyuanVideo 1.5
BenchmarkSora 2HunyuanVideo 1.5Source
LMArena Text to Video
Higher rating1,340.1Rank 16, 95% interval 1,333.5 to 1,346.7, 62,692 votes
1,169.2Rank 35, 95% interval 1,152.9 to 1,185.5, 4,290 votes
LMArenaRetrieved 2026-08-26

Higher rating is better. A clear lead is shown only when the published 95% intervals do not overlap.

Key to the marks:better valuebetter, but not conclusivetie

Visual comparison

Sample videos

Only runs from both exact model versions under the same archived prompt and first-output rule appear here. Provider showcases and substitute versions remain excluded.

Controlled practical comparison. Both exact model versions received the same archived prompt in one request. The first output counts. Playback uses the first five seconds, 1280 × 720 pixels, 30 fps, no crop, and muted audio. Provider-side randomness and prompt processing remain model characteristics.

Motion and physical consistency

Anatomy, object consistency, camera tracking, rain, and reflections

Sora 2

Run details

fal.ai. fal-ai/sora-2/text-to-video. 16:9, one request, first output, requested 8 s, seed provider random, request {"prompt":"Generate exactly one continuous five-second 16:9 cinematic video. A fictional bicycle courier in a mustard-yellow raincoat rides a red city bicycle from left to right across a rain-wet Hamburg street at blue hour for the entire clip. The camera tracks her smoothly at waist height. Both hands remain visible on the handlebars, the bicycle geometry stays physically correct, and a transparent umbrella fixed to the bicycle moves naturally in the wind. Wet-street reflections and raindrops follow realistic physics. No cuts, no slow motion, no text, no logos, no subtitles, and no audio.","resolution":"720p","aspect_ratio":"16:9","duration":8}, configured price estimate $0.80.

Generated Aug 24, 2026. Source

HunyuanVideo 1.5

Run details

fal.ai. fal-ai/hunyuan-video-v1.5/text-to-video. 16:9, one request, first output, requested 5 s, seed 1928510983, request {"prompt":"Generate exactly one continuous five-second 16:9 cinematic video. A fictional bicycle courier in a mustard-yellow raincoat rides a red city bicycle from left to right across a rain-wet Hamburg street at blue hour for the entire clip. The camera tracks her smoothly at waist height. Both hands remain visible on the handlebars, the bicycle geometry stays physically correct, and a transparent umbrella fixed to the bicycle moves naturally in the wind. Wet-street reflections and raindrops follow realistic physics. No cuts, no slow motion, no text, no logos, no subtitles, and no audio.","aspect_ratio":"16:9","resolution":"480p","num_frames":121,"num_inference_steps":28,"enable_prompt_expansion":false}, configured price estimate $0.38.

Generated Aug 24, 2026. Source

Prompt and integrity hashes

Generate exactly one continuous five-second 16:9 cinematic video. A fictional bicycle courier in a mustard-yellow raincoat rides a red city bicycle from left to right across a rain-wet Hamburg street at blue hour for the entire clip. The camera tracks her smoothly at waist height. Both hands remain visible on the handlebars, the bicycle geometry stays physically correct, and a transparent umbrella fixed to the bicycle moves naturally in the wind. Wet-street reflections and raindrops follow realistic physics. No cuts, no slow motion, no text, no logos, no subtitles, and no audio.

Prompt SHA 256: 206a9182584e93790fd6df608e8a2065fe2c8f747e85aeaaf4c8ea84040e6299

Sora 2, Original SHA 256: 017b48183183b34289b75843b0d07bf5ba373800adfcc4d148a17979ef322538

Sora 2, Normalized SHA 256: 0b391c3b44e83bcd4deed191bbd65ca924385d2a274a42ff95d79722023c0728

HunyuanVideo 1.5, Original SHA 256: 79196c1316e097a8236eb7d97ae43945a29780e837c02e65807fcf67443596dd

HunyuanVideo 1.5, Normalized SHA 256: bbf97d2e087a0d0a77e5f8f254ff74eae8808e8449597e6581e946410c445538

How to read the data

Preference is not the same as production fit

A higher LMArena rating means that outputs from this exact configuration were preferred more often in blind comparisons.

Motion control, identity consistency, audio, turnaround time, moderation, and licensing may matter more than the aggregate preference rating.

Prices are representative public minute rates, not a normalized cost benchmark.

More matchups

Related video model comparisons

Compare Sora 2 and HunyuanVideo 1.5 with other important video models using the same benchmark logic.

View the complete video model comparison