Reference

Frameworks & tools

This is the perishable companion to Choosing a framework, which owns the durable decision logic: when a framework earns its keep, when the raw SDK wins. This page is the current landscape—what's shipping, who's mature, who's in maintenance mode. Reviewed quarterly; every star count and version carries an as-of date.

Cards are grouped by job-to-be-done, not vendor. Within each group, the first card is the one we'd reach for by default; the rest are the cases where it's wrong.


Agent & workflow frameworks

LangGraph

Low-level stateful agent runtime from LangChain Inc. Execution modeled as a directed graph: nodes are Python functions, edges are routing rules, a checkpointer serializes state after every node so the run survives a process restart.

Latestlanggraph==1.2.4 (June 2, 2026)
Stars34,458 (as of June 2026)
Pick it whenLong-running, interruptible workflows; agents with complex conditional routing; teams already on LangSmith for tracing.
Skip it whenYour workflow fits in a single function, or you need provider-agnostic tracing (LangGraph couples tightly to LangSmith).
Production usersUber, LinkedIn, Klarna (named in the v1.0 announcement).
Docsdocs.langchain.com

PydanticAI

Type-safe agent framework from the Pydantic team. Agents are generic over result type and dependency type; the framework auto-retries tool calls when the LLM returns invalid data, using Pydantic validation directly.

Latestv1.107.0 (June 10, 2026)
Stars17,700 (as of June 2026)
Pick it whenTeams already using Pydantic; single-agent or light multi-agent setups; apps where structured-output reliability is primary.
Skip it whenYou need durable checkpointing out of the box (no PostgresSaver equivalent), or you need the largest community ecosystem.
Docsai.pydantic.dev

CrewAI

Standalone multi-agent framework (no LangChain dependency since v1.0). Agents modeled as role-playing participants with role/goal/backstory. Two primitives: Crews (autonomous collaboration) and Flows (event-driven pipelines).

Latest1.14.7 (June 11, 2026); 201 releases, weekly cadence
Stars53,300 (as of June 2026)
Pick it whenRole-based multi-agent workflows where personas matter; event-driven pipelines with conditional branching.
Skip it whenYou need fine-grained durable state, or the task is simple enough that role/goal/backstory is just prompt overhead.
Docsdocs.crewai.com

AutoGen (Microsoft)

Multi-agent conversation framework. Original repo (microsoft/autogen, 58,900 stars) is now in maintenance mode as of early 2026. Active development moved to microsoft/agent-framework. The new framework drops the conversation metaphor for event-driven orchestration.

StatusAutoGen 0.4 — maintenance mode. New: Agent Framework (preview)
Stars58,900 (legacy repo, as of June 2026)
Pick it whenOnly if you have an existing AutoGen deployment. For greenfield, evaluate Agent Framework or LangGraph instead.
Skip it whenStarting fresh (maintenance mode means declining community investment).
Docsmicrosoft.github.io/autogen

DSPy

Prompt optimization framework from Stanford NLP. You write programs with typed signatures; DSPy compiles them into optimized prompts/few-shot sets via teleprompters (optimizers) that search over prompt strategies.

Latest3.2.1 (June 2026)
Stars35,000 (as of June 2026)
Pick it whenYou have a labeled eval set and want automatic prompt/few-shot optimization; pipelines with multiple cascading LLM calls that need joint tuning.
Skip it whenYou don't have eval data to optimize against, or your system has one simple prompt that doesn't justify the abstraction overhead.
Docsdspy.ai

RAG & data frameworks

LlamaIndex

Data framework for LLM applications. Owns the indexing-to-query pipeline: document loaders, chunking strategies, vector store integrations, query engines, and retrieval abstractions. The most mature RAG-specific toolkit.

Latestllama-index-core==0.14.22 (June 2026)
Stars50,100 (as of June 2026)
Pick it whenComplex document ingestion pipelines; multi-source RAG with metadata routing; you need pre-built integrations for 160+ data sources.
Skip it whenSimple single-source retrieval where pgvector + 50 lines of code does the job, or you want to avoid the large dependency tree.
Docsdocs.llamaindex.ai

Infrastructure

LiteLLM

Unified API gateway for 100+ LLM providers. One completion() call, swap providers by changing the model string. Adds load balancing, fallbacks, spend tracking, and rate limiting across all providers.

Latestv1.88.1 (June 2026)
Stars50,200 (as of June 2026)
Pick it whenMulti-provider routing, fallbacks, or spend tracking; you want OpenAI-compatible interface to Anthropic/Gemini/open-weight models.
Skip it whenSingle-provider app with no routing needs; the extra abstraction layer adds latency (~2-5ms) you can't afford on voice paths.
Docsdocs.litellm.ai

vLLM

High-throughput LLM serving engine. PagedAttention for efficient KV cache management, continuous batching, tensor/pipeline parallelism. The default self-hosting inference server in 2026.

Latestv0.8.x (June 2026)
Stars~50,000 (as of June 2026)
Pick it whenSelf-hosting open-weight models; you need continuous batching, speculative decoding, or multi-GPU inference.
Skip it whenUsing hosted APIs exclusively; your traffic is too low to justify GPU ops overhead.
Docsdocs.vllm.ai

Voice & realtime

LiveKit Agents

Real-time voice/video AI agent framework built on the LiveKit WebRTC infrastructure. Handles VAD, STT, LLM orchestration, TTS, and turn-taking as a pipeline with pluggable components.

Latestv1.6.x (mid-2026)
Stars~4,200 (agents repo, as of early 2026)
Pick it whenProduction voice agents with WebRTC transport; you need infrastructure (TURN, rooms, recording) alongside the AI pipeline.
Skip it whenText-only agents, or you're using a native speech-to-speech model (OpenAI Realtime API) that bypasses the cascaded pipeline entirely.
Docsdocs.livekit.io/agents

Pipecat

Open-source framework for building voice and multimodal AI agents. Frame-based pipeline architecture: audio/video frames flow through processors (VAD, STT, LLM, TTS) that you compose declaratively.

Latestv0.0.55 (early 2026)
Stars~12,000 (as of early 2026)
Pick it whenMultimodal agents (voice + vision); you want a lighter, more composable alternative to LiveKit's full infrastructure stack.
Skip it whenYou need production-grade WebRTC rooms, recording, and TURN servers out of the box (LiveKit owns that layer).
Docsdocs.pipecat.ai

Observability

Langfuse

Open-source LLM observability platform. Tracing, prompt management, evaluation, and cost tracking with full OpenTelemetry support. Self-hostable or cloud.

Latestv3.185.0 (June 2026)
Stars29,000 (as of June 2026)
Pick it whenYou want open-source observability you can self-host; OTel-native tracing matters; budget-conscious teams.
Skip it whenYour team is already deep in LangSmith and unwilling to migrate; you need enterprise SSO/RBAC without self-hosting.
Docslangfuse.com/docs

LangSmith

Closed-source observability and evaluation platform from LangChain Inc. Tightly integrated with LangGraph and the LangChain ecosystem. Strongest eval/annotation tooling for teams in that ecosystem.

LatestCloud SaaS (continuously deployed)
Pick it whenYou're already using LangGraph/LangChain; you want the tightest integration with those tools; enterprise procurement is not an issue.
Skip it whenYou need self-hosting, OTel-native export, or provider-agnostic tooling. Vendor lock-in is real.
Docsdocs.smith.langchain.com

Phoenix (Arize)

Open-source LLM observability tool focused on tracing, evaluation, and experimentation. OTel-native with OpenInference conventions. Lightweight alternative to the full-platform plays.

Latestv10.x (June 2026)
Stars10,100 (as of June 2026)
Pick it whenLocal-first development; you want fast trace visualization without standing up infrastructure; already using Arize for ML monitoring.
Skip it whenYou need a full production platform with prompt management, annotation queues, and team collaboration (Langfuse or LangSmith). Note: ELv2 license, not true open-source.
Docsdocs.arize.com/phoenix

Browser automation

browser-use

AI-native browser agent framework. Connects LLMs to a browser via DOM extraction and structured actions. The model sees a cleaned DOM tree and issues high-level commands (click, type, navigate).

Latestv0.13.1 (June 2026)
Stars98,400 (as of June 2026)
Pick it whenBuilding AI agents that need to browse the web; you want DOM-driven actions (faster, cheaper than screenshots).
Skip it whenThe target site is behind auth you control (use their API instead), or you need screenshot-driven universality (use Computer Use).
Docsdocs.browser-use.com

Playwright

Cross-browser automation library from Microsoft. Not AI-specific, but the default execution substrate under AI browser agents. Headless Chromium/Firefox/WebKit with auto-waiting, network interception, and codegen.

Latestv1.52.x (early 2026)
Stars~89,800 (as of early 2026)
Pick it whenYou need the browser execution layer under browser-use or your own agent; E2E testing of AI-powered UIs; web scraping at scale.
Skip it whenYou're using Computer Use (screenshot-driven, no DOM access needed), or the task doesn't involve a browser at all.
Docsplaywright.dev/python

Last reviewed: June 12, 2026. Star counts and versions shift weekly. If a number here is more than 90 days old, check the linked docs. The decision logic for when to use a framework at all lives in Chapter 7.12.