Technical / Code-First Frameworks · Updated 1 August 2026
Top Frameworks
Ranked by real-world demand. As of 1 August 2026, LangChain is the most-sought framework — index 82.2 of 100.
Framework finder
Which agent framework should you use?
There are 25 on the board. Answer 3 quick questions and get the 2–3 that actually fit how you build — not just the most popular.
How these frameworks differ
The board ranks by demand. This map places each tool by low-level / composable → high-level / opinionated and single-agent → multi-agent team, so you can choose by fit, not just popularity.
| # | Tool | Index | Demand | GitHub ⭐ | Searches/mo |
|---|---|---|---|---|---|
| 1 | LangChainThe most widely adopted LLM application toolkit — chains, tools, and hundreds of integrations in Python and JavaScript. | 82.2 | 185k | 111k | |
| 2 | LangGraphGraph-based agent orchestration from the LangChain team — explicit nodes, edges, and state for controllable, inspectable workflows. | 67.2 | 42k | 33k | |
| 3 | LlamaIndexThe data framework for LLM apps — connects agents to your own documents with best-in-class retrieval (RAG). | 65.9 | 54k | 13k | |
| 4 | CrewAIRole-based multi-agent framework — define a crew of cooperating agents and let them divide the work. | 61.9 | 56k | 13k | |
| 5 | HaystackDeepset's production-oriented framework for search, RAG, and agent pipelines. | 58.2 | 26k | 4.3k | |
| 6 | Spring AIBrings LLM and agent patterns to the Spring ecosystem for Java teams. | 57.2 | 9.2k | 4.4k | |
| 7 | Microsoft Agent FrameworkMicrosoft's unified agent framework for Python and .NET — the successor to AutoGen and Semantic Kernel, with graph-based workflows and enterprise plumbing. | 56.6 | 13k | 6.6k | |
| 8 | OpenAI Agents SDKOpenAI's lightweight SDK for building agents, with built-in handoffs, guardrails, and tracing. | 52.8 | 32k | 5.4k | |
| 9 | Google ADKGoogle's Agent Development Kit — Gemini-native, multi-agent by design, wired into Google Cloud. | 52.8 | 21k | 9.9k | |
| 10 | Anthropic Claude Agent SDK~Anthropic's SDK for building agents on Claude — the same harness that powers Claude Code. | 52.2 | 9.4k | 9.9k | |
| 11 | Strands AgentsAWS's model-driven agent SDK — a few lines of Python from prototype to production. | 51.5 | 6.8k | 51k | |
| 12 | DSPyStanford's framework for programming rather than prompting LLMs — it optimises prompts and pipelines automatically. | 50.7 | 37k | 5.8k | |
| 13 | Semantic KernelMicrosoft's enterprise orchestration framework for .NET, Python, and Java — now converging into Microsoft Agent Framework. | 50.3 | 29k | 2.4k | |
| 14 | MastraTypeScript-first agent framework — workflows, RAG, and evals with a strong developer experience. | 46.9 | 27k | 1.6k | |
| 15 | Pydantic AIType-safe agent framework from the Pydantic team — structured, validated outputs built in. | 46.5 | 19k | — | |
| 16 | Letta (MemGPT)Agents with persistent, self-editing memory — grown out of the MemGPT research — that learn across sessions. | 44.0 | 24k | 1.9k | |
| 17 | Eliza (elizaOS)TypeScript framework for autonomous character agents, popular for social and web3 personas. | 43.9 | 19k | 720 | |
| 18 | AgnoFast, lightweight framework for multimodal, multi-agent systems with built-in memory and knowledge. | 42.8 | 42k | 2.7k | |
| 19 | CAMEL~Research-driven multi-agent framework for building and studying communicating, role-playing agent societies. | 40.7 | 18k | 2.5k | |
| 20 | smolagentsHugging Face's minimal agent library — agents that think in code, in about a thousand lines. | 38.7 | 29k | 1.6k | |
| 21 | Atomic AgentsModular, schema-first framework for composing predictable agent pipelines from small parts. | 37.2 | 6.1k | 390 | |
| 22 | GriptapePython framework for secure, structured agent workflows with off-prompt data handling. | 36.1 | 2.6k | 1.4k | |
| 23 | OpenAI SwarmOpenAI's experimental, educational framework for lightweight multi-agent handoffs — superseded by the Agents SDK. | 36.0 | 22k | 480 | |
| 24 | Guidance~Constrained-generation library that steers model output with grammars and templates instead of free text. | 32.0 | 22k | 20 | |
| 25 | Marvin~Prefect's lightweight toolkit for turning LLM calls into typed, reliable Python functions. | 31.5 | 6.2k | 456 |
Index = market demand (search + AI demand) weighted most heavily, then adoption, real usage, momentum, production-readiness and community. ◇ proprietary / closed-source (no public repo — scored on demand). ~ demand estimated from product-site traffic. Methodology → · All boards →