Announcements

Stay informed about Portkey's newest releases, features, and improvements

  1. ⚡ Claude Sonnet 4.5 is live on Portkey!

    Anthropic’s latest Claude Sonnet 4.5 is now available through the Portkey AI Gateway.

    With Portkey, you can bring Claude Sonnet 4.5 into production with:

    ✅ Unified access via the AI Gateway

    ✅ Full observability: logs, latency, and spend

    ✅ Budget and rate-limit controls across teams and apps

    claude sonnet 4.5

    2

  2. Regex replace guardrail ✏️

    Your data policies aren’t always one-size-fits-all.

    Alongside our PII guardrail, we’ve introduced a Regex Replace Guardrail for specific redactions.

    Define custom regex patterns and automatically mask sensitive data before it reaches the model so that you can:

    • Mask sensitive data (e.g., emails, phone numbers)
    • Normalize inputs before they reach the model
    • Enforce consistent formatting across requests
    regex-replace-guardrails

    Read more here -> https://lnkd.in/dgkPGrQF

  3. Metadata-based model access guardrail 🛡️

    We’ve introduced a new guardrail that lets you restrict model access based on metadata key–value pairs at runtime.

    This means you can:

    • Enforce tenant, region, or environment-based restrictions dynamically
    • Ensure compliance policies are checked per request
    • Control model usage without changing your app logic

    Granular governance, enforced in real time.

    2

  4. Happening today: LLMs in Prod, San Francisco!

    We’re teaming up with Exa to host the next edition of LLMs in Prod. Expect talks, panels, and conversations with PG&E, Postman, Palo Alto Networks, LinkedIn, Cerebras, and more, focused on running LLMs in production at scale.

    📍 Exa HQ, San Francisco

    📅 Today, 5–8 PM

    Are we seeing you today? (few spots left!)

  5. Unified Count Tokens Endpoint!

    We’ve introduced a single endpoint for counting tokens across AWS Bedrock, Vertex AI, and Anthropic.

    This makes it easier to:

    • Get an estimate of tokens before sending the request
    • Manage rate limits and costs
    • Enforce routing and quota rules inside your app
    • Optimize prompts to be a specific length

    Read more here ->

  6. Unified finish_reason parameter across providers ✅

    By default, finish_reason parameter is now mapped to an OpenAI-compatible value for consistency. If you want to keep the original provider-returned value, set x-portkey-strict-openai-compliance = false.

  7. Conditional Router now supports request parameters ⚡

    Metadata-based routing works well when you have policy-driven or static strategies, for example, routing by tenant, region, or business tier. But what if you want to optimize routing per request?

    Portkey now allows you to set up conditional routing on request parameters too.

    This gives you more flexibility and better performance, letting you route dynamically based on each request’s conditions (like input length, urgency, or task type).

    👉 See how you can set it up here →

  8. Quick updates to improve your experience!

    We’ve rolled out a set of improvements across multiple providers 👇

    • DashScope: Updated supported parameters
    • Vertex AI: Added timeRangeFilter support for Google Search tool
    • OpenAI & Azure OpenAI: New parameters added for GPT-5 compatibility
    • OpenRouter: Now returns reasoning messages when provided by the model
    • Fireworks:
    1. Better handling of non-ASCII characters in file uploads
    2. Removed unnecessary response transforms → faster processing