FluiqFluiqDocs

Pillars

  • SecurityBlock attacks, redact PII and secrets
  • ObservabilityTrace every call, cost, and latency
  • EvaluationScore responses and whole agent runs

Part of Evaluation

  • DatasetsGolden sets that capture whole agent runs
  • Prompt ManagementVersion and deploy prompt templates

Across the platform

  • AlertsPush eval and security events to Slack

LLM Providers

  • OpenAI
  • Anthropic
  • Google Gemini
  • Google Vertex AI

Agent Frameworks

  • LangChain
  • LangGraph
  • CrewAI
  • Google ADK
  • MCP

Vector Databases

  • Pinecone
  • Chroma
  • Weaviate
  • FAISS
  • Qdrant
14 integrations · zero wrappersView all
Pricing

Learn

  • BlogsWriting on evals, security & cost
  • FAQPricing, evals, security & data

Build

  • Fluiq DocsGuides, concepts & SDK reference
  • Code SamplesCopy-paste integration snippets

Tools

  • Response Gate DemoWhat an LLM leaks while refusing
  • LLM Cost CalculatorCompare OpenAI, Claude & Gemini
  • polygateOpen-source unified LLM client
  • InfragerDiagrams to secure Terraform
Contact
Get API key

Language

Getting Started

Quickstart

Pillars

SecurityObservabilityEvaluation

Part of Evaluation

DatasetsPrompts

Reference

ConfigurationAlerts
Code Examples
SDK reference

Fluiq Python SDK

Two lines of Python instrument any AI agent or LLM pipeline. Auto-traced integrations for OpenAI, Anthropic, Gemini, LangChain, and MCP, plus a @trace decorator for everything else. Server-side security scanning, Redis caching, and LLM-as-judge evaluations are one method call each, all on Fluiq infrastructure, nothing to deploy.

Quickstart

Install the package, grab an API key, and call instrument() once at startup. Every supported LLM call from that point is traced automatically.

1

Install

Python
pip install fluiq
2

Get an API key

Create a free account and copy your key from the dashboard. Sign up →

3

Instrument once at startup

Python
import fluiq

fluiq.instrument(api_key="fl_...") # or set FLUIQ_API_KEY to environment

# Every OpenAI / Anthropic / Gemini / LangChain / MCP call from this
# point on is traced automatically, free and unlimited on every tier.
# Opt in to more when you want it:
fluiq.eval(thresholds={"hallucination": 0.8})  # LLM-as-judge scoring
fluiq.secure()     # Security scanning (all plans)

Next steps

Open the dashboard
Watch traces stream in, inspect costs by node, and review quality scores.
Go to dashboard
Add the eval gate
Gate every PR on hallucination, faithfulness, and relevancy thresholds.
See eval gate setup
Build a golden dataset
Pin whole agent runs from production and batch-evaluate them before you ship a change.
See dataset runs
Manage prompts
Discover, edit, evaluate, and promote prompt templates to dev, staging, and production.
Open Prompts