Create your API key
Create an account and generate a key from the Velokey console.
sk-••••••••••••7f3a
API KEYACTIVE
Text, image, and video models. One API.
Access Top AI Models
in One API
Access leading text, image, and video models through one API and save 20–80% on costs. Switch models without re-integrating and pay only for what you use.
Get $0.5 in free credits at sign-up — about 20 free AI images, never expires
/ 02 - MODELS
One API for text, image, and video models
Explore models by modality, capability, and price. Start with the right model today and switch when your requirements change.
Explore Video APIs
Need a faster model route.
Router picked GPT fallback.
message...
Featured models
Seedance / Kling / Veo / Wan
Pick your model
Powerful AI models, simple API access.
Seedance 2.0
Cinematic multi-modal video generation
- doubao-seedance-2-0-text-to-video
- doubao-seedance-2-0-image-to-video
- doubao-seedance-2-0-reference-to-video
Kling V3
High-fidelity video, image & omni generation
- kling-v3-video-generation
- kling-v3-image-generation
- kling-v3-omni-video-generation
GPT Image 2
State-of-the-art image generation & editing
- gpt-image-2
- gpt-image-2_1k
Sora 2
OpenAI's cinematic text-to-video
- sora-2
Nano Banana Pro
Google's fast conversational image model
- gemini-3-pro-image-preview
- gemini-3.1-flash-image-preview
Wan 2.7
Alibaba's multi-modal video & image suite
- wan2.7-t2v
- wan2.7-i2v
- wan2.7-image
Seedream 5.0
ByteDance's quality-first text-to-image
- doubao-seedream-5-0
- doubao-seedream-5-0-pro
- doubao-seedream-4-5
Grok Imagine
xAI's imaginative image & video generation
- grok-imagine-image
- grok-imagine-video
- grok-imagine-video-1.5-preview
Veo 3.1
Google's cinematic video with native audio
- veo-3.1-generate-preview
- veo-3.1-fast-generate-preview
- veo-3.1-lite-generate-preview
PixVerse V6
Fast stylized video generation
- pixverse-v6-t2v
- pixverse-v6-it2v
- pixverse-v6-r2v
Vidu Q3
High-speed multi-reference video generation
- viduq3-pro_text2video
- viduq3-turbo_text2video
- viduq3-pro_img2video
Qwen Image 2.0
Bilingual text rendering, precise editing
- qwen-image-2.0
- qwen-image-2.0-pro
Switch Models with One Familiar API
Keep your existing OpenAI-compatible client. Update your Base URL and API key, then choose the model you want to call.
velokey - OpenAI-compatible quickstart
from openai import OpenAI
client = OpenAI(api_key="YOUR_VELOKEY_API_KEY", base_url="https://api.velokey.ai/v1")
response = client.chat.completions.create(model="YOUR_MODEL_ID", messages=[...])
print(response.choices[0].message.content)From API Key to First Call in Minutes
If you already use an OpenAI-compatible SDK, migration starts with a Base URL and API key change.
Update your Base URL
Keep your existing SDK and point it to the Velokey endpoint.
-api.openai.com/v1
+api.velokey.ai/v1
BASE URL1 LINE CHANGED
Choose a model and build
Call text, image, and video models through one account and interface.
GPT-5.5ClaudeGeminiDeepSeekSeedanceKlingVeoWanGPT ImageNano BananaQwenGrok
Compare models before you commit
Evaluate models by capability, context, speed, billing unit, and price before adding them to your workflow.
GPT API family
Compare GPT models across reasoning, coding, context and cost for general intelligence tasks from prototype to production.
For reference only
GPT-5.6GPQA—SWE-bench96.2%VLMArena—Input—Output—
GPT-5.5GPQA93.6%SWE-bench58.6%PLMArena1475Input—Output—
GPT-5.4GPQA92.8%SWE-bench57.7%PLMArena1467Input—Output—
GPT-5.4 miniGPQA—SWE-bench54.4%PLMArena1449Input—Output—
Reliable access. Transparent operation.
Understand how requests are routed, what each call costs, and what happens when an upstream route becomes unavailable.
Smart model routing
One request, automatically routed to the faster, more stable endpoint available.
request → velokey.ai router
latency check
cost optimization
automatic failover
response ← best provider
Automatic failover
When multiple routes are available, failed requests can move to a healthy fallback route.
REQ
Provider ATimeout
Provider BHealthy
FAILOVERAUTO
One console, full control
Track request status, token usage, latency, errors, and spend from one place.
Last 30-day spend
93.78Credits
Daily spend trend
52 Credits
Transparent model pricing
See token, image, and video pricing before you call. Pay only for actual usage.
Claude Sonnet 4.6$3.00/ 1M tokens
GPT Image 2$0.006/ image
Seedance 2.0$0.29/ second
usage-based meteringpay only for actual usage
/ 07 - FAQ
Common questions
Key details about pricing, migration, model access, reliability, and data handling on Velokey.
Velokey is a unified AI API platform for accessing leading text, image, and video models through one account, API key, and integration.
Velokey uses pay-as-you-go pricing. Language models are billed by token usage, while image and video billing units depend on the selected model. Prices are visible before you call.
If your app already uses an OpenAI-compatible SDK, migration starts by changing the Base URL and API key. You can then choose a supported model without replacing the rest of your client.
Velokey focuses on leading language, image, and video models. The model catalog shows currently available models, versions, capabilities, and pricing.
When multiple healthy routes are available and failover is enabled, Velokey can move requests to a fallback route based on availability, latency, and cost.
Velokey does not retain prompt or model-output content and does not use customer API data for training. Limited operational metadata is retained for billing, security, troubleshooting, and support.
Your next call starts here
One API, one bill, every major model.
gpt-5.4-mini
qwen3.6-plus
claude-fable-5claude-sonnet-5
doubao-seed-2-1-prodoubao-seed-2-1-turbodoubao-seed-evolvingclaude-opus-5claude-opus-4-8
gemini-3.1-pro-previewgemini-3-pro-previewgemini-3.1-flash-lite-previewgemini-3-flash-previewgemini-3.5-flash
deepseek-v4-flashdeepseek-v4-pro
ernie-5.0ernie-5.1
grok-4.20grok-4.3
kimi-k2.5kimi-k2.6kimi-k2.7-codekimi-k3
MiniMax-M2.7MiniMax-M3qwen3.6-flashqwen3.7-plusqwen3.7-maxgpt-5.4gpt-5.5claude-opus-4-7claude-opus-4-6claude-opus-4-5-20251101claude-haiku-4-5-20251001claude-sonnet-4-6claude-sonnet-4-5-20250929
glm-5.2glm-5glm-5.1wan2.7-imagesora-2gpt-image-2qwen-image-2.0doubao-seedream-5-0doubao-seedream-4-5seedance-2.0wan2.7pixverse-v6Nano Banana Progrok-imagine-imagegrok-imagine-videogrok-imagine-video-1.5-previewhappyhorse-1.1
kling-v3-videokling-v3-imagekling-v3-omni-videokling-v3-omni-imageNano Banana 2omni-flash
pixverse-c1qwen-image-2.0-proveo-3.1-generate-preview
viduq3-providuq3-turbo