Portkey – AI Gateway & LLMOps Control Plane for Multi-Model Apps | Quasa.io

#Quasa #QUA #Portkey

Portkey is an AI gateway and LLMOps control plane that sits between your app and the models you call. One OpenAI-compatible API routes traffic across 1,600+ models and 40+ providers, with fallbacks, load balancing, retries, caching, logs, traces, cost tracking, guardrails, prompt management, and an MCP gateway. The core gateway is MIT open source; the hosted product adds observability and governance. Commercial path is now tied to Palo Alto Networks / Prisma AIRS. Core tools: Universal API, routing & failover, simple + semantic cache, observability (logs/traces/cost), guardrails & PII redaction, prompt studio, virtual keys/budgets, MCP gateway. Pricing: Developer free (10k logs/mo, 3-day retention). Production $49/mo (100k logs, +$9 per extra 100k). Enterprise custom (VPC, SSO, SOC2/HIPAA). OSS self-host is free software plus your infra.

𝐂𝗢𝗥𝗘 𝗦𝗧𝗥𝗘𝗡𝗚𝗧𝗛𝗦
• One API across a very wide model catalog
• Production controls: fallbacks, cache, budgets, guardrails
• Deep request observability
• Free tier + MIT open-source gateway
• Enterprise path via Prisma AIRS  

𝗜𝗗𝗘𝗔𝗟 𝗙𝗢𝗥
Teams running multi-model LLM apps or agents who need reliability, cost control, and governance without wiring every provider separately.

𝗛𝗜𝗚𝗛𝗟𝗜𝗚𝗛𝗧𝗦
• Drop-in OpenAI-compatible integration
• Semantic caching can cut token spend
• Prompt versioning + playground
• 3,000+ GenAI teams claimed  

𝗣𝗢𝗧𝗘𝗡𝗧𝗜𝗔𝗟 𝗖𝗢𝗡𝗦𝗜𝗗𝗘𝗥𝗔𝗧𝗜𝗢𝗡𝗦
• Paid value is mostly logs/retention and governance, not raw throughput
• Free logs stop at 10k (requests still flow)
• Acquisition changes the long-term roadmap
• Not a model marketplace like OpenRouter  

𝗢𝗩𝗘𝗥𝗔𝗟𝗟 𝗩𝗘𝗥𝗗𝗜𝗖𝗧
4.4/5 stars. Strong production gateway if you already bring your own keys and need routing plus observability. Start on the free tier, then move to $49 only when you need longer logs, semantic cache, and RBAC. Earn 1 QUA reward by reviewing on Quasa.io too!

𝗚𝗘𝗧 𝗦𝗧𝗔𝗥𝗧𝗘𝗗: https://quasa.io/projects/portkey