AI Gateway

Datawizz’s AI platform gives you full control over AI request routing, protects against abuse, and boosts reliability with fallbacks and caching. Manage everything from the dashboard - no code changes needed - keeping app logic separate from AI infrastructure.

Smart Routing

Smart Routing

Smart Routing

Route requests to different models based on tags, metadata, prompt size or end-use

Datawizz gives you the flexibility to deploy mission specific models and route requests to the right one every time. Keep your code simple by making a single AI request, and abstract away intricate routing logic.

Datawizz gives you the flexibility to deploy mission specific models and route requests to the right one every time. Keep your code simple by making a single AI request, and abstract away intricate routing logic.

Datawizz gives you the flexibility to deploy mission specific models and route requests to the right one every time. Keep your code simple by making a single AI request, and abstract away intricate routing logic.

Tag based routing

Deploy different models to handle 
different tasks or data types, 
and route based on tags

Tag based routing

Deploy different models to handle 
different tasks or data types, 
and route based on tags

Tag based routing

Deploy different models to handle 
different tasks or data types, 
and route based on tags

Size based routing

Deploy models based on context size - giving more optimized models for smaller requests while supporting bigger contexts

Size based routing

Deploy models based on context size - giving more optimized models for smaller requests while supporting bigger contexts

Size based routing

Deploy models based on context size - giving more optimized models for smaller requests while supporting bigger contexts

User based routing

Serve different models to users based 
on subscription tiers, data residency location or privacy requirements

User based routing

Serve different models to users based 
on subscription tiers, data residency location or privacy requirements

User based routing

Serve different models to users based 
on subscription tiers, data residency location or privacy requirements

Multi-provider orchestration

Multi-provider orchestration

Multi-provider orchestration

Datawizz makes leveraging multiple AI model providers a breeze

All models are normalized to a standard input and output format - abstracting away model preferences from application logic. This flexibility allows you to swap models, route between multiple providers or create multi-provider fallbacks without adding unnecessary complexity to your code.

Your App

Your App

Your App

Large Language Model

Fallback Model

Your App

Large Language Model

Fallback Model

Your App

Large Language Model

Fallback Model

Your App

Provider Fallbacks

Provider Fallbacks

Provider Fallbacks

Define fallbacks in case
of an outage

Datawizz lets you define fallbacks, so in case one AI provider has an outage, requests get seamlessly re-routed to a different prodders. This lets you increase app reliability despite provider outages.

Semantic Caching

Semantic Caching

Semantic Caching

Cache responses and serve them based on direct or semantic similarity

Save money on repeated requests. Datawizz gives you fine-grained control over similarity tolerances and comparison functions to make sure users get the right response every time.

Guardrails

Guardrails

Guardrails

Configure guardrails

and policies to control inputs and outputs to your model

Save money on repeated requests. Datawizz gives you fine-grained control over similarity tolerances and comparison functions to make sure users get the right response every time.

Anonymize

Scan and reduct PII and other sensitive information from AI prompts.

Ban Code

Detect and block code snippets in your prompts.

Ban Topics

Prevent your prompts from discussing sensitive / undesirable topics.

Prompt Injection

Detect and prevent prompt-injection based attacks.

Globally Distributed

Globally Distributed

Globally Distributed

Stellar reliability and deployment in hundreds of outposts around the world

Datawizz runs on the global Cloudflare platform. This network places your logic and AI models closer to your users, reducing latency and increasing application performance.

Pricing

Pricing

Pricing

Get most of the Datawizz platform

By default routing and logging requests through the Datawizz platform is free. Some features require the Pro tier, which is billed per request.

By default routing and logging requests through the Datawizz platform is free. Some features require the Pro tier, which is billed per request.

By default routing and logging requests through the Datawizz platform is free. Some features require the Pro tier, which is billed per request.

Basic

Free

  • Record Requests

  • Log Human Feedback

  • Basic Model Routing

  • Log Retention - 90 days

Pro

$1

/ 10,000 requests

  • Record Requests

  • Log Human Feedback

  • Basic Model Routing

  • Smart Model Routing

  • Fallbacks

  • Caching

  • AI Guard Policies

  • Basic Model Routing

  • Log Retention - Forever

Contact Us

Contact Us

Contact Us