GOAMLDXQCNKS
Text, images, and audio
Generate anything through a single interface. Every major model in one place, with the same request shape.
Browse all modelsThe unified interface
for every model.
Better prices, better uptime, no subscriptions.
No credit card · Free starting credit · Cancel anytime
POST /v1/chat/completions200 OK · 812 ms
Requestmodel: auto
Summarise this Q3 financial report into 5 points, then flag the biggest risk.
policy: balancedmax_tokens: 4096
input1.284 tok
output96 tok
cost$0.0027
Routing3 candidates
Achosen
claude-sonnet-4-5
Anthropic · $0.0027
92
O
gpt-5.2
OpenAI · $0.0081
88
G
gemini-3-pro
Google · $0.0064
85
Failover active. If one provider fails, the request moves to another before the first token is sent.
700T+
Tokens / month
10M+
Active users
64+
Providers
500+
Models
Why AIGet
Built for teams that run AI in production
One integration, the whole model ecosystem. No vendor lock-in.
aiget/auto
GOA
Higher availability
Reliable AI models via distributed infrastructure. Fall back to another provider automatically when one goes down.
How failover worksLatency
Price and performance
Keep costs in check without sacrificing speed. The gateway runs at the Singapore edge, close to Asia.
See pricingCustom data policies
Prompts are never stored or used for training. You decide which providers may receive your data.
Read the policyFeatured models
500+ active models across 64 providers
How it works
Three steps, under ten minutes.
1Sign up
Create an account to get started. You can set up an organization and invite your team later.
+18
2Buy credits
Top up with QRIS, GoPay, or virtual account. Credits never expire and work with any model.
QRISRp 100.000
3Get your API key
Create an API key and start making requests. Fully OpenAI compatible.
sk-aiget-••••7f3a
Integration
Change one line, not your whole codebase
The OpenAI SDK, LangChain, the Vercel AI SDK, and any agent framework work without meaningful changes.
- Full SSE streaming
- Tool calling and JSON mode
- Swap models without deploying
- Cross-provider fallback
curl https://api.aiget.tech/v1/chat/completions \
-H "Authorization: Bearer $AIGET_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "aiget/auto",
"messages": [
{ "role": "user", "content": "Summarise this report" }
],
"stream": true
}'Observability
Cost you can account for
Every request is recorded: input tokens, output tokens, the model used, and the actual cost. No invoice you cannot explain.
- p95 latency
- 1,840 ms
- Gateway overhead
- < 80 ms
- 90-day uptime
- 99.98%
- Average TTFT
- 286 ms
Daily cost$1.47 total
Oct 1Oct 25
Pricing
Pay for what you actually use
No markup on token prices. One platform fee, stated plainly.
Hobby
Free
For experiments and side projects.
Start for free- Starting credit
- 3 API keys
- 60 requests/min
- 7-day logs
- Community support
Most popular
Pro
$29/month
For products already in production.
Start 14-day trial- No platform markup
- Unlimited API keys
- 600 requests/min
- 90-day logs + export
- Priority failover
- Email support
Enterprise
Custom
For teams with compliance needs.
Talk to us- BYOK and dedicated capacity
- SSO / SAML and RBAC
- 99.99% SLA
- VPC / on-prem deploy
- Zero retention
- Slack support
Frequently asked questions
What is an AI router?
A layer between your application and AI model providers. You send one request to one endpoint; the router picks the model, handles retries, and records cost. Your app never needs to know which provider was used.
Is it compatible with the OpenAI SDK?
Yes. Requests and responses follow the Chat Completions format, including SSE streaming, tool calling, and JSON mode. Just change the base URL and API key.
How does the router choose a model?
You set the policy: cost, quality, latency, or balanced. The router filters candidates by context window and capabilities, then picks the highest score.
Are my prompts stored?
No. Prompt and response content is not stored by default. Only metadata is recorded: tokens, cost, latency, and model.
What happens when credits run out?
Requests are rejected with a 402 before reaching the provider. Enable auto top-up so your service never stops unexpectedly.
Can I use my own provider keys?
Yes. The provider bills you directly, and AIGet only adds a small platform fee.
Ready to move your AI onto one endpoint?
Create your first API key today. See how much you can save.