Maxim AI API

aid: maxim:maxim-api
name: Maxim AI API
description: The Maxim REST API provides programmatic access to prompts, workflows, agents, datasets,
  evaluators, test runs, logging, tracing, models, alerts, and log repositories. It supports OpenTelemetry-compatible
  ingestion and is paired with Python and TypeScript SDKs.
humanURL: https://www.getmaxim.ai/docs
baseURL: https://api.getmaxim.ai
tags:
- LLM
- Agents
- Evaluation
- Tracing
- Observability
- Prompts
- Datasets
properties:
- type: Documentation
  url: https://www.getmaxim.ai/docs
- type: GettingStarted
  url: https://www.getmaxim.ai/docs/introduction/quickstart/quickstart
- type: Signup
  url: https://app.getmaxim.ai/sign-up
- type: SDKs
  url: https://github.com/maximhq/maxim-py
- type: SDKs
  url: https://github.com/maximhq/maxim-js
- type: GitHubRepository
  url: https://github.com/maximhq/maxim-cookbooks
- type: Pricing
  url: https://www.getmaxim.ai/pricing
- type: Authentication
  url: https://www.getmaxim.ai/docs/introduction/quickstart/quickstart
features:
- name: Distributed Tracing
  description: Capture spans, generations, retrievals, tool calls, and sessions across multi-step agent
    runs.
- name: 50+ Pre-Built Evaluators
  description: Use AI-powered, programmatic, statistical, and voice-based evaluators out of the box, or
    build custom ones via SDK.
- name: Offline and Online Evaluations
  description: Run dataset-driven test suites pre-deploy and apply auto-evaluation to live logs in production.
- name: No-Code Agent Builder
  description: Visually compose multi-agent workflows with conditional logic, loops, and HTTP endpoint
    agents.
- name: Prompt Management
  description: Version prompts, manage partials, and roll out conditional deployments with automated optimization.
- name: Datasets and Synthetic Data
  description: Curate datasets, attach files, generate synthetic samples, and manage splits and columns.
- name: Voice Agent Simulation
  description: Simulate voice agent conversations and evaluate them with voice-aware metrics.
- name: Real-Time Dashboards and Alerts
  description: Build custom dashboards over logs and trigger Slack and PagerDuty alerts on quality regressions.
- name: OpenTelemetry Support
  description: Ingest spans via OTLP from any OTEL-instrumented service.
- name: Self-Hosted and Hybrid Deployments
  description: Run Maxim in your own VPC with zero-touch or data-plane peering modes.
useCases:
- name: Agent Quality Engineering
  description: Evaluate, simulate, and monitor production agents across releases.
- name: Prompt Optimization
  description: A/B test and tune prompts against curated datasets.
- name: Voice AI Observability
  description: Trace and evaluate voice agents built on ElevenLabs, OpenAI Realtime, and LiveKit.
- name: RAG Evaluation
  description: Measure retrieval relevance and groundedness on large datasets.
- name: Compliance and Audit
  description: Enterprise audit logging and human annotation workflows for regulated industries.
integrations:
- name: OpenAI
- name: Anthropic
- name: Google Gemini
- name: Mistral
- name: Groq
- name: Together AI
- name: Fireworks
- name: xAI Grok
- name: Cerebras
- name: OpenRouter
- name: LangChain
- name: LangGraph
- name: CrewAI
- name: LlamaIndex
- name: Pydantic AI
- name: Vercel AI SDK
- name: AWS Bedrock
- name: LiteLLM
- name: Portkey
- name: ElevenLabs
- name: LiveKit
- name: Google ADK
- name: OpenTelemetry
- name: Slack
- name: PagerDuty
authentication:
- type: API Key
  description: Maxim API keys are generated in the workspace settings and passed on each request; SAML
    SSO (Google, Okta) and 2FA are available for dashboard access.