HunyuanVideo 1.5 API Pricing

Tencent's lightweight 8.3B-parameter open video model generating up to 121 frames (~5 seconds) with strong motion coherence and text interpretation at 480p or 720p. Optimized for fast, cost-efficient inference on consumer-grade GPUs; available via fal, WaveSpeed, Novita, and other resellers with per-second billing. Tracked across 2 providers · 5 price points · billed per-second — all normalized to cost per second.

$0.040/second
cheapest verified rate · WaveSpeedAI · standard tier · 720p · audio off · checked Jul 30, 2026

Try WaveSpeedAI Full cheapest ranking

HunyuanVideo 1.5 pricing — quick facts

  • Cheapest verified rate: $0.040/second on WaveSpeedAI (standard, 720p, audio off) — checked Jul 30, 2026.
  • Across 2 providers, HunyuanVideo 1.5 runs $0.020–$0.075/second (3.8× spread).
  • A 5-second clip costs $0.20 at the cheapest verified rate, up to $0.38 at the priciest tracked tier.

Bottom line: WaveSpeedAI at $0.040/s is both the cheapest verified and the sensible default choice.

Every HunyuanVideo 1.5 price, by provider

Sorted by normalized $/second. Hover a price for its derivation; every row links to the provider's published source. Range: $0.020$0.075/s (3.8× spread).

ProviderTrustModeTierResAudio$ / second5s clipVerifiedSource
WaveSpeedAICHEAPEST VERIFIEDTry →mediumText → Videostandard480p$0.020$0.10verified Jul 30, 2026source ↗
WaveSpeedAImediumText → Videostandard720p$0.040$0.20verified Jul 30, 2026source ↗
WaveSpeedAImediumImage → Videostandard720p$0.040$0.20verified Jul 30, 2026source ↗
fal.aihighText → Videostandard480p$0.075$0.38verified Jul 30, 2026source ↗
fal.aihighImage → Videostandard480p$0.075$0.38verified Jul 30, 2026source ↗
How each price was normalized (derivations & sources)
  • WaveSpeedAI · standard · 480p: WaveSpeed HunyuanVideo 1.5 t2v page: 480p = $0.02/s (cheapest published Hunyuan 1.5 rate anywhere) source ↗
  • WaveSpeedAI · standard · 720p: WaveSpeed HunyuanVideo 1.5 t2v page: 720p = $0.04/s (480p = $0.02/s); cheapest verified 720p Hunyuan rate source ↗
  • WaveSpeedAI · standard · 720p: WaveSpeed HunyuanVideo 1.5 i2v page: 720p = $0.04/s (480p = $0.02/s) source ↗
  • fal.ai · standard · 480p: fal HunyuanVideo 1.5 t2v page: $0.075/s at 480p (fal endpoint is 480p only, up to 121 frames) source ↗
  • fal.ai · standard · 480p: fal HunyuanVideo 1.5 i2v page: $0.075/s (480p output) source ↗

What a clip actually costs

At the cheapest verified rate vs the most expensive tracked rate.

Clip lengthCheapest (WaveSpeedAI)Most expensive trackedYou save
5 seconds$0.20$0.38$0.17
10 seconds$0.40$0.75$0.35
15 seconds$0.60$1.13$0.53

Estimate your own volume with the cost calculator.

Choosing a HunyuanVideo 1.5 provider

Computed from the live dataset — cheapest first. “+” is why you'd pick it, “−” is why you'd skip it.

WaveSpeedAI medium

from $0.040/s · standard · 720p
  • Cheapest verified rate we track: $0.040/s
  • Simple per-second billing

fal.ai high

from $0.075/s · standard · 480p
  • Verified $0.075/s at 480p
  • High-trust production choice
  • Simple per-second billing
  • 1.9× the cheapest verified rate

HunyuanVideo 1.5 pricing FAQ

How much does the HunyuanVideo 1.5 API cost?
HunyuanVideo 1.5 API pricing ranges from $0.020 to $0.075 per second of generated video depending on provider, resolution and tier — a 3.8x spread. The cheapest verified 720p rate is $0.040/second on WaveSpeedAI.
What is the cheapest HunyuanVideo 1.5 API provider?
As of July 2026, the cheapest verified HunyuanVideo 1.5 rate at 720p is $0.040 per second on WaveSpeedAI (standard tier, audio off).
How much does a 5-second HunyuanVideo 1.5 video cost?
At the cheapest verified rate ($0.040/s on WaveSpeedAI), a 5-second clip costs about $0.20. At the most expensive tracked rate it would cost $0.38 — provider choice matters more than most developers expect.
How is HunyuanVideo 1.5 priced?
Across the 5 price points we track, HunyuanVideo 1.5 is billed per second of generated video. We normalize every scheme to USD per second of output video so providers can be compared directly. Full derivations are shown per row and on our methodology page.

Compare with other models