Cheapest AI Video Generation API in 2026 — Price Breakdown | GenRelay
You're building a feature that generates video clips, and you need to choose a model before you can write a billing estimate. The problem: AI video pricing is listed in different units across different vendors. Some charge per second of output video. Others charge a flat fee per generation regardless of clip length. Comparing them head-to-head requires converting everything to the same denominator.
This article gives you the real numbers for every video model available on GenRelay in August 2026, converts them to a common cost-per-second basis where applicable, and works through typical workloads so you know what each model will actually cost at volume.
What AI video generation models are available, and how are they billed?
GenRelay provides access to four video models with two distinct billing structures — per second of output and per generation.
| Model | Billing | 720p | 1080p | 4K |
|---|---|---|---|---|
| Veo 3.1 Lite | Per second | $0.060/s | $0.120/s | $0.180/s |
| Grok Imagine 1.5 | Per second | $0.022/s | $0.022/s | — |
| Grok Imagine 1.0 | Per second | $0.010/s | $0.010/s | — |
| Omni Flash | Per generation | $0.10/clip | $0.15/clip | — |
Key differences:
- Veo 3.1 Lite bills per second of video output with resolution-tiered pricing. A 10-second clip costs 10× the per-second rate.
- Grok Imagine 1.0 and 1.5 bill per second at a flat rate across resolutions. Grok 1.0 is text-to-video; Grok 1.5 handles image-to-video (i2v).
- Omni Flash bills per generation at a fixed price regardless of clip duration — useful for workloads where you generate fixed-length clips and want predictable unit costs.
How does per-second billing compare to per-generation billing?
Per-second billing is transparent when you control clip length — you know the cost before you submit the request. Per-generation billing gives you a fixed price per clip regardless of length, which simplifies cost prediction if your clips have variable duration.
Which is cheaper depends on clip length. For very short clips (2–3 seconds), per-generation can be more expensive on a per-second basis. For longer clips where the output fills a fixed-price slot, per-generation becomes more favorable.
To make the comparison concrete, here is what a 5-second clip costs across all models at 720p:
| Model | Billing unit | Cost for a 5-second 720p clip |
|---|---|---|
| Grok Imagine 1.0 | $0.010/s | $0.050 |
| Omni Flash | $0.10/clip | $0.10 |
| Grok Imagine 1.5 (i2v) | $0.022/s | $0.110 |
| Veo 3.1 Lite | $0.060/s | $0.30 |
At 5 seconds and 720p, Grok Imagine 1.0 is the lowest-cost option at $0.050 per clip. Omni Flash at $0.10 is competitive for this clip length and becomes more favorable as clips get longer (since the price stays fixed).
What does 100 clips per month cost with each model?
The table below shows the monthly cost for 100 clips at two representative clip lengths. This covers the most common production workload range for apps that offer video generation on demand.
Monthly cost — 100 clips at 720p
| Model | 5-second clips | 15-second clips |
|---|---|---|
| Grok Imagine 1.0 | $5.00 | $15.00 |
| Omni Flash | $10.00 | $10.00 |
| Grok Imagine 1.5 | $11.00 | $33.00 |
| Veo 3.1 Lite | $30.00 | $90.00 |
At 5-second clip length, Grok 1.0 is clearly the most affordable option. At 15 seconds, Omni Flash's flat rate ($10.00 total) beats Grok 1.0 ($15.00) because the fixed price no longer scales with duration.
Monthly cost — 100 clips at 1080p
| Model | 5-second clips | 15-second clips |
|---|---|---|
| Grok Imagine 1.0 | $5.00 | $15.00 |
| Omni Flash | $15.00 | $15.00 |
| Grok Imagine 1.5 | $11.00 | $33.00 |
| Veo 3.1 Lite | $60.00 | $180.00 |
At 1080p, Grok 1.0's per-second rate remains flat while Veo 3.1 Lite doubles its 720p rate. For high-resolution workloads where quality can be Grok-grade, Grok 1.0 stays the most affordable.
What model is the most affordable for short-form social content?
For short-form clips (5–8 seconds at 720p), Grok Imagine 1.0 at $0.010/s gives the lowest per-clip cost: $0.05–$0.08 per clip. At 1,000 clips per month, that is $50–$80 — a monthly spend that is viable even for early-stage apps with free tiers.
Grok 1.0 is text-to-video. If your use case starts from a reference image — product photo to animated clip, portrait to motion video — Grok 1.5 at $0.022/s adds image-to-video input at roughly 2.2× the Grok 1.0 rate.
For more detail on the Grok Imagine models and their generation parameters, see the Grok video API guide.
How does resolution choice affect your monthly video API bill?
Resolution choice has a significant cost multiplier depending on the model.
Veo 3.1 Lite: resolution doubles the price at each tier. Moving from 720p to 1080p doubles cost ($0.060 → $0.120/s). Moving to 4K adds another 50% ($0.120 → $0.180/s). For a 10,000-second monthly output volume, that gap is $600/month (720p) vs. $1,200/month (1080p) vs. $1,800/month (4K).
Grok Imagine 1.0 and 1.5: current pricing on GenRelay is flat across resolutions. The same $0.010/s or $0.022/s applies at 720p and 1080p.
Omni Flash: two tiers — $0.10 at 720p and $0.15 at 1080p. The 50% premium for 1080p is fixed regardless of clip length, making the resolution decision straightforward.
Practical guidance: for most product integrations that target social platforms, 720p is sufficient for web delivery and keeps costs at their lowest. Reserve 1080p for use cases where output is viewed on large screens or downloaded as a final deliverable.
For a deeper look at how resolution affects the Veo 3.1 billing model specifically, see the Veo 3 API pricing breakdown.
When does it make sense to use Veo 3.1 Lite despite the higher cost?
Veo 3.1 Lite generates noticeably higher-quality video output than Grok 1.0, including more coherent motion, better prompt adherence, and support for text-to-video with audio synthesis. The cost premium ($0.060/s vs. $0.010/s) is 6× at 720p, but the quality gap can justify it in specific contexts:
- Marketing asset production: one-off hero clips where quality matters and generation frequency is low.
- Paid creative tools: if your users pay per credit and expect premium output, the cost is passed through.
- Workflows combining t2v + i2v + ref2v: Veo 3.1 supports all three generation modes including reference-to-video. Grok 1.0 is text-to-video only.
For workloads where output volume is high and quality requirements are moderate — background loops, social filler content, draft storyboards — Grok 1.0 at $0.010/s or Omni Flash at $0.10/clip are the cost-efficient paths.
How do you call the GenRelay video API in Python?
All video models share the same endpoint. Switch models by changing the model field.
import os
import time
import requests
API_BASE = "https://genrelay.ai/v1"
HEADERS = {"Authorization": f"Bearer {os.environ['GENRELAY_API_KEY']}"}
def generate_video(prompt: str, model: str = "grok-1.0", duration: int = 5, resolution: str = "720p") -> str:
"""Submit a video generation job and poll until complete. Returns video URL."""
# Submit job
r = requests.post(
f"{API_BASE}/videos/generations",
headers=HEADERS,
json={"model": model, "prompt": prompt, "duration": duration, "resolution": resolution},
timeout=30,
)
r.raise_for_status()
job_id = r.json()["id"]
# Poll for completion
for _ in range(60): # up to ~5 minutes
time.sleep(5)
status_r = requests.get(f"{API_BASE}/videos/generations/{job_id}", headers=HEADERS)
status_r.raise_for_status()
job = status_r.json()
if job["status"] == "succeeded":
return job["output"]["url"]
if job["status"] == "failed":
raise RuntimeError(f"Video generation failed: {job.get('error')}")
raise TimeoutError("Video generation did not complete within timeout")
Switch to Grok 1.5 for image-to-video by adding an image_url parameter:
def generate_video_from_image(image_url: str, prompt: str, duration: int = 5) -> str:
r = requests.post(
f"{API_BASE}/videos/generations",
headers=HEADERS,
json={
"model": "grok-1.5",
"prompt": prompt,
"image_url": image_url,
"duration": duration,
"resolution": "720p",
},
timeout=30,
)
r.raise_for_status()
job_id = r.json()["id"]
# poll as above
return poll_job(job_id)
Frequently asked questions
Which model is the cheapest per second of video output?
Grok Imagine 1.0 at $0.010/s is the lowest per-second rate on GenRelay as of August 2026. At 720p, a 10-second clip costs $0.10 total.
Does Veo 3.1 Lite support audio generation?
Yes. Veo 3.1 includes audio synthesis as part of its video generation pipeline. Grok 1.0 and Grok 1.5 are video-only (silent output). Omni Flash does not include audio.
Is there a free tier for video generation?
GenRelay includes free credits on signup that apply to video models. The credits are sufficient for testing generation calls and building your integration before committing to a paid plan.
What happens if the video generation job fails?
The job status returns "failed" with an error field. Video generation can fail due to prompt policy violations, upstream model errors, or timeout. Retry with a modified prompt or after a short delay. For detailed async polling patterns, see the async video API polling guide.
Does GenRelay charge for failed generation attempts?
GenRelay does not charge for jobs that fail before producing output. Charges apply only when a video is successfully generated.