Google Veo 3
GOOGLE:VEO-3-GENMultimodal VideoBusyLoad:88%SLA:99.85%排队 3.5s
DeepMind Native Multimodal Generative Video Network
Base Rate
75 Credits / run
24h +8.12%
DeepMind native audio-visual video model mastering cinematic framing, focus pulling, and fluid temporal coherence.
24h Total Calls
0.89Mcalls
+8.12% vs yesterday
Total Generations
0.58Mvids
+7.4% · Cinema Camera 88%
Avg Latency (P50)
12.4s
8s Cinema 12.4s · Camera Paths
Availability SLA
99.85%
Failover <120ms
Model Telemetry & Volume Chart
GOOGLE:VEO-3-GENInvocations
38,955次+20.53%
Live Telemetry
Operational Performance
Production-grade Latency & Concurrency
P50 TTFT420 ms
P99 Latency18900 ms
Throughput TPS
Peak Concurrency480 req
Packet Loss Rate0.001%
Cost & Billing Matrix
Credits and Token Breakdown
Base Task Cost75 Credits
Input Cost / 1k Tokens¥0.025
Output Cost / 1k Tokens¥0.080
4K Upscale Multiplier2.2x
VIP Discount至高 20% 优惠
Technical Specifications
Parameters, Context & Resolutions
ArchitectureDeepMind Native Multimodal Generative Video Network
Parameters24 Billion
Context Window4,096 tokens
Max Resolution1080P Native (4K Cinema Post)
Max Duration8s Extended
Global Edge Gateways
Live Region Latency & Load Balancing
East China (Shanghai)
30ms
South China (Guangzhou)
34ms
Hong Kong Gateway
38ms
Asia Pacific (Singapore)
50ms
US West (Silicon Valley)
122ms
EU Central (Frankfurt)
148ms
Comparable Models in Same Domain
Models with similar creative modalities
Kling 2.6 Pro (可灵)
KUAISHOU:KLING-2.6-PRO
Kuaishou state-of-the-art video model with cinematic camera paths, high-motion human dynamics, and physical consistency.
Rate60 Credits / run
Hailuo 2.3 (海螺视频)
MINIMAX:HAILUO-2.3
MiniMax cinematic text/image-to-video flagship renowned for subtle facial expressions and complex physical world dynamics.
Rate55 Credits / run