Skip to content
Stay informed about Portkey's newest releases, features, and improvements
After this date, the model will no longer be available or operable. This also impacts fine-tuned models built on top of this base model.
If you’re using Mistral-large-2411 in production, it’s a good time to review deployments and plan a migration to supported alternatives.
Google is deprecating Gemini 2.0 Flash and Gemini 2.0 Flash Lite on March 31, 2026 across Gemini API and Google AI Studio.
If your apps or agents rely on these models, you’ll need to migrate to supported alternatives like Gemini 2.5 Flash or 2.5 Flash Lite before that date.
MCP adoption is accelerating rapidly, with servers and tools now being used across teams, environments, and agent workflows.
As MCP usage scales, teams start running into a consistent set of challenges:
Portkey’s MCP Gateway solves this!
It is a control layer to securely connect MCP servers, standardize authentication, manage tool access, and get clear visibility into MCP usage, without changing existing agents or MCP servers.
Check it out ->
Keeping LLM pricing accurate is difficult because it isn’t static. Models change, pricing shifts, and new billing dimensions are introduced over time.
To handle this, we built a never-goes-old database that powers cost calculations ~$250,000 in LLM spend, every day.
We’ve now open-sourced the pricing database produced by this system so others can inspect how pricing is tracked, reuse the data, and understand the assumptions behind cost calculations.
The database covers 2,000+ models across 40+ providers, including thinking tokens, cache pricing, context tiers, and non-token fees. It’s available via a free public API with no authentication required.
Take a look ->https://portkey.ai/models
Running OpenCode in shared or production environments?
By using Portkey alongside OpenCode, teams can add the controls needed to run it at scale, without changing how developers work.
You get:
✅ Access control, budgets, and usage limits at the platform layer
✅ Centralized governance without modifying OpenCode itself
✅ Full observability across usage, latency, and spend
✅ Guardrails to keep usage predictable in shared environments
See how you can run OpenCode with governance
3
You can now access Anthropic models on Azure via the native /messages endpoint, so you can keep the standard Anthropic request format while getting:
✅ Unified access through the Portkey Gateway
✅ Guardrails for safe, compliant usage
✅ Full observability across logs, latency, and spend
✅ Team-level budgets and rate-limit controls
2
You can now start using LLMs hosted on Oracle Cloud Infrastructure (OCI) in production with:
✅ Support for Cohere, Meta and Llama models via OCI Generative AI
✅ Guardrails for safe, compliant usage
✅ Full observability across logs, latency, and spend
✅ Team-level budgets and rate-limit controls
See how you can add Oracle models to Portkey.
Gemini 3 Flash delivers fast reasoning and multimodal understanding at lower latency and cost compared to larger models, while still handling complex text, image, audio, and video tasks.
With Portkey, you can bring this model into production with:
✅ Unified access via the Gateway
✅ Guardrails for safe, compliant usage
✅ Full observability across logs, latency, and spend
✅ Team-level budgets and rate-limit controls
Try it today!
Because AgentCore supports OpenAI-compatible frameworks, you can integrate Portkey without modifying your agent code while keeping AgentCore’s runtime, gateway, and memory services intact.
With this setup, you get:
See how you can implement it here.
Sticky load balancing ensures that requests sharing the same identifier are consistently routed to the same target.
This is useful for:
Read more about this here ->
