deepset-ai /haystack
Open-source AI orchestration framework for building context-engineered, production-ready LLM applications. Design modular pipelines and agent workflows with explicit control over retrieval, routing, memory, and generation. Built for scalable agents, RAG, multimodal applications, semantic search, and conversational systems.
★ 27,010⑂ 3,469👁 164⑃ 7,907● 4,683▤ 256
+4
日增长
+89
周增长
27,010
stars
3,469
forks
164
watchers
7,907
prs
4,683
issues
256
releases
README
| | | | ------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | CI/CD | | | Docs | | | Package | !PyPI - Downloads !PyPI - Python Version | | Meta | | Haystack is an open-source AI orchestration framework for building production-ready LLM applications in Python. Design modular pipelines and agent workflows with explicit control over retrieval, routing, memory, and generation. Build scalable RAG systems, multimodal applications, semantic search, question answering, and autonomous agents, all in a transparent architecture that lets you expe
Top issues
#11039RFC: Signed receipts for Haystack pipeline component calls20 comments · 2026-04-05#10810EU AI Act compliance scan results — Haystack scored #1, wanted to validate findings17 comments · 2026-03-12#11836feat: Pipeline run recording & deterministic replay - turn every run into a shareable, diffable, testable artifact10 comments · 2026-06-30#10973Feature: Post-RAG FactChecker Pipeline Component & Header-Aware Document Splitter9 comments · 2026-03-29#10821Feature: GuardrailProvider interface for automated tool-call policy enforcement9 comments · 2026-03-14#11266Proposal: Transaction Protocol for idempotent, auditable agent pipelines7 comments · 2026-05-06#6970Streamlit Stream Callback in Haystack 2.x7 comments · 2024-02-11#12060Allow Human-in-the-Loop to review fully materialized tool calls6 comments · 2026-07-18
Discussions
#13705mcp-slim: Zero-dependency MCP stdio proxy for Haystack MCPToolset (-50% to -98% schema tokens & SecretGuard)2 comments · 2026-11-12#13626OpenAIChatGenerator against a non-OpenAI base URL: api_base_url vs OPENAI_BASE_URL, stacked retries, and keeping keys out of pipeline YAML0 comments · 2026-11-09#13609Testing the scope of remembered tool approval: ARQUIA-EVAL-001 open fixture0 comments · 2026-11-07#13584ArtiGram: a free gallery for AI-made interactive artifacts, with a remote MCP server0 comments · 2026-11-06#13407Testing error propagation from MCPToolset through a two-component Haystack pipeline0 comments · 2026-10-23#13205A Haystack component that prepares a payment request instead of paying: agent proposes, human signs0 comments · 2026-10-10#13048Per-run dollar budgets for Haystack pipelines and agents (no custom component)0 comments · 2026-10-01#12969Your retrieval budget may be based on a context window you are not actually being served1 comments · 2026-09-26
Latest release
v3.5.0 v3.5.0 2026-11-10⭐️ Highlights 🧭 Classify and Route with the OpenAI Decisions API Three new components use the OpenAI Decisions API to choose how to handle your data without instructions. OpenAIDecisionsTextRouter sends text to a named output per label, OpenAIDecisionsImageRouter does the same for images and can send low-confidence decisions to a fallback branch, and OpenAIDecisionsDocumentClassifier answers choice, score, and yes/no questions about documents and stores the answers in their metadata. python from haystack.components.routers import OpenAIDecisionsTextRouter router = OpenAIDecisionsTextRouter