dbotmakerthe decision bot

Machine readout · Software & AI

AI Agent Framework Statistics 2026

This page collects figures on the open source frameworks developers use to build AI agents, with a focus on LangChain and LangGraph, CrewAI, Microsoft AutoGen and LlamaIndex, plus newer entrants such as the OpenAI Agents SDK, Vercel AI SDK, Pydantic AI and Mastra. It covers developer survey shares from Stack Overflow, JetBrains and IBM, GitHub stars, PyPI and npm download counts, company-reported enterprise adoption, and the Model Context Protocol that links agents to tools. Survey data is global and runs from 2024 to 2026; repository and download counts were read on November 13, 2026.

Figures
53
Publishers
36
Updated
Nov 13, 2026

Quotable findings

By the DBotMaker research desk

  1. Finding 012026
    30%

    30.2% of Stack Overflow's 2026 survey respondents who build with AI agents had used LangChain for agent orchestration in the past year, the highest share of any framework.

    Source: Stack Overflow

  2. Finding 022026
    157,203,220

    The langchain package on PyPI was downloaded 157,203,220 times in the month to November 12, 2026.

    Source: PyPI Stats

  3. Finding 032026
    149.5k

    The langchain-ai/langchain repository had 149.5k GitHub stars in November 2026, the most of any agent framework covered here.

    Source: GitHub

  4. Finding 042026
    139,881,777

    Vercel's ai package, the AI SDK for TypeScript, was downloaded 139,881,777 times from npm in October 2026.

    Source: npm

  5. Finding 052026
    4.5%

    4.5% of agent builders in Stack Overflow's 2026 survey had used Microsoft's AutoGen, below the 12% recorded in 2025.

    Source: Stack Overflow

  6. Finding 062026
    65%

    CrewAI said in November 2026 that 65% of the Fortune 500 used its platform, up from nearly half in October 2024.

    Source: CrewAI

  7. Finding 072025
    57.3%

    57.3% of 1,340 professionals surveyed by LangChain in late 2025 had agents running in production, and another 30.4% were developing agents with plans to deploy.

    Source: LangChain

  8. Finding 082025
    97M+

    Official MCP SDKs had more than 97 million monthly downloads across Python and TypeScript in December 2025.

    Source: Anthropic

  9. On file53figures in 7 channels, each with the exact source sentence.Browse allCSV

THE BOT'S SHORT ANSWER

LangChain is the most used AI agent framework: 30% of agent builders in Stack Overflow's 2026 survey used it, its langchain package was downloaded 157.2 million times on PyPI in the month to November 12, 2026, and its main repository has 149.5k GitHub stars. CrewAI, AutoGen and LlamaIndex each have 50,000 to 62,000 stars but far lower download counts, from 2.6 million (CrewAI) to 0.4 million (AutoGen) PyPI downloads a month.

53 FIGURES · 36 PUBLISHERS · NOVEMBER 13, 2026

Reading the numbers

LangChain leads on every public measure. Stack Overflow's 2026 survey found 30% of agent builders had used it, its main GitHub repository holds 149.5k stars, and PyPI Stats counted 157.2 million downloads of the langchain package in a single month. LangGraph, its orchestration layer, adds another 58.7 million PyPI downloads and 44.9k stars. The company itself reports more than 350 million monthly open source downloads across all its packages.

The second tier is crowded and the order depends on the measure. On GitHub, AutoGen (61.8k stars), CrewAI (60.9k) and LlamaIndex (52.9k) are close. On PyPI, CrewAI (2.6 million monthly downloads) and llama-index (2.0 million) are well ahead of autogen-agentchat (0.4 million). In Stack Overflow's 2026 survey, the OpenAI Agents SDK (24%) and LangGraph (19%) sit well above CrewAI (6%) and AutoGen (5%), and the OpenAI SDK's 19 million monthly PyPI downloads point the same way.

Enterprise adoption figures come mostly from the vendors. LangChain says 35% of the Fortune 500 use its services and it has more than 7,000 LangSmith customers; CrewAI says 65% of the Fortune 500 use its platform and that it runs around 450 million agents a month. These numbers are useful for direction but are not audited, and their definitions are not published. LangChain's own survey found 57.3% of respondents had agents in production, while Stack Overflow's broader 2025 sample found 52% of developers did not use agents at all.

In TypeScript, the picture is different: Vercel's AI SDK drew 139.9 million npm downloads in October 2026, about five times @langchain/core, and Mastra and the OpenAI Agents SDK each approach 10 million a month. Across all languages, the Model Context Protocol, with more than 10,000 public servers and 97 million monthly SDK downloads by December 2025, has become a shared layer that most frameworks support.

Facts the bot knows

CH 01Which agent frameworks developers use

9 figures
  1. #00130%

    30.2% of Stack Overflow's 2026 survey respondents who build with AI agents had used LangChain for agent orchestration in the past year, the highest share of any framework.

    Stack Overflow, Stack Overflow Developer Survey 2026: AI agent orchestration tools (Jul 2026)

    Exact wording at the source
    LangChain leads at 30%, followed by OpenAI Agents SDK at 24%.
  2. #00223.8%

    23.8% of agent builders in Stack Overflow's 2026 survey had used the OpenAI Agents SDK, making it the second most used orchestration tool.

    Stack Overflow, Stack Overflow Developer Survey 2026: AI agent orchestration tools (Jul 2026)

    Exact wording at the source
    OpenAI Agents SDK 1,048 23.8%
  3. #00318.8%

    18.8% of agent builders in Stack Overflow's 2026 survey had used LangGraph, LangChain's graph-based agent runtime.

    Stack Overflow, Stack Overflow Developer Survey 2026: AI agent orchestration tools (Jul 2026)

    Exact wording at the source
    LangGraph 825 18.8%
  4. #0045.5%

    5.5% of agent builders in Stack Overflow's 2026 survey had used CrewAI in the past year, down from 7.5% in the 2025 survey.

    Stack Overflow, Stack Overflow Developer Survey 2026: AI agent orchestration tools (Jul 2026)

    Exact wording at the source
    CrewAI 242 5.5%
  5. #0054.5%

    4.5% of agent builders in Stack Overflow's 2026 survey had used Microsoft's AutoGen, below the 12% recorded in 2025.

    Stack Overflow, Stack Overflow Developer Survey 2026: AI agent orchestration tools (Jul 2026)

    Exact wording at the source
    AutoGen 200 4.5%
  6. #00633%

    33% of developers building agents told Stack Overflow in 2025 that they used LangChain, second only to the local model runner Ollama at 51%.

    Stack Overflow, 2025 Stack Overflow Developer Survey: AI (Jul 2025)

    Exact wording at the source
    Among developers building agents, Ollama (51%) and LangChain (33%) are the most-used frameworks.
  7. #00713.3%

    13.3% of agent builders in Stack Overflow's 2025 survey had used LlamaIndex for agent orchestration, ahead of AutoGen at 12% and CrewAI at 7.5%.

    Stack Overflow, 2025 Stack Overflow Developer Survey: AI (Jul 2025)

    Exact wording at the source
    Llama Index 13.3% AutoGen (Microsoft) 12% Zapier 11.8% CrewAI 7.5%
  8. #00899%

    99% of 1,000 enterprise AI application developers surveyed by IBM and Morning Consult in 2025 said they were exploring or developing AI agents.

    IBM, AI Agents in 2025: Expectations vs. Reality (Mar 2025)

    Exact wording at the source
    IBM and Morning Consult did a survey of 1,000 developers who are building AI applications for enterprise, and 99% of them said they are exploring or developing AI agents
  9. #00962%

    62% of 24,534 developers in JetBrains' 2025 survey relied on at least one AI coding assistant, agent or AI code editor, and 85% regularly used AI tools for development.

    JetBrains, The State of Developer Ecosystem 2025 (Oct 2025)

    Exact wording at the source
    85% of developers regularly use AI tools for coding and development, and 62% rely on at least one AI coding assistant, agent, or code editor.

Chart

Agent orchestration tools used in the past year, 2025

Share of Stack Overflow 2025 survey respondents who use or develop AI agents; multiple answers allowed.

  • Ollama51.1%
  • LangChain32.9%
  • LangGraph16.2%
  • Vertex AI15.1%
  • Amazon Bedrock Agents14.5%
  • LlamaIndex13.3%
  • AutoGen12%
  • CrewAI7.5%

Source: Stack Overflow, 2025 Stack Overflow Developer Survey: AI

CH 02GitHub stars and open source activity

8 figures
  1. #010149.5k

    The langchain-ai/langchain repository had 149.5k GitHub stars in November 2026, the most of any agent framework covered here.

    GitHub, langchain-ai/langchain repository on GitHub (Nov 2026)

    Exact wording at the source
    Stars 149.5k stars
  2. #01161.8k

    Microsoft's AutoGen repository had 61.8k GitHub stars in November 2026, even though Microsoft now directs new feature work to Microsoft Agent Framework.

    GitHub, microsoft/autogen repository on GitHub (Nov 2026)

    Exact wording at the source
    Stars 61.8k stars
  3. #01260.9k

    The crewAIInc/crewAI repository had 60.9k GitHub stars and 9.2k forks in November 2026.

    GitHub, crewAIInc/crewAI repository on GitHub (Nov 2026)

    Exact wording at the source
    Stars 60.9k stars Watchers 389 watching Forks 9.2k forks
  4. #01352.9k

    The run-llama/llama_index repository had 52.9k GitHub stars in November 2026.

    GitHub, run-llama/llama_index repository on GitHub (Nov 2026)

    Exact wording at the source
    Stars 52.9k stars
  5. #01444.9k

    LangGraph, LangChain's agent orchestration library, had 44.9k GitHub stars in November 2026.

    GitHub, langchain-ai/langgraph repository on GitHub (Nov 2026)

    Exact wording at the source
    Stars 44.9k stars
  6. #01540,000

    LlamaIndex reported almost 40,000 GitHub stars in March 2025, so its count grew by roughly a third over the following 20 months.

    LlamaIndex, Announcing our Series A and LlamaCloud general availability (Mar 2025)

    Exact wording at the source
    We've got more than 3 million monthly downloads, almost 40,000 GitHub stars, and hundreds of thousands of active developers worldwide.
  7. #0161.1 million

    More than 1.1 million public GitHub repositories used an LLM SDK by August 2025, and 693,867 of them were created in the prior 12 months.

    GitHub, Octoverse: A new developer joins GitHub every second as AI leads TypeScript to #1 (Oct 2025)

    Exact wording at the source
    More than 1.1 million public repositories now use an LLM SDK with 693,867 of these projects created in just the past 12 months alone
  8. #0174.3 million

    AI-related repositories on GitHub exceeded 4.3 million in 2025, nearly double the count of less than two years earlier, according to GitHub's Octoverse report.

    GitHub, Octoverse: A new developer joins GitHub every second as AI leads TypeScript to #1 (Oct 2025)

    Exact wording at the source
    AI-related repositories on GitHub now exceed 4.3 million, nearly doubling in less than two years.

CH 03PyPI downloads for Python agent frameworks

11 figures
  1. #018157,203,220

    The langchain package on PyPI was downloaded 157,203,220 times in the month to November 12, 2026.

    PyPI Stats, Recent downloads for langchain (Nov 2026)

    Exact wording at the source
    "last_month":157203220
  2. #01958,666,621

    The langgraph package on PyPI was downloaded 58,666,621 times in the month to November 12, 2026.

    PyPI Stats, Recent downloads for langgraph (Nov 2026)

    Exact wording at the source
    "last_month":58666621
  3. #02018,956,914

    OpenAI's openai-agents package on PyPI was downloaded 18,956,914 times in the month to November 12, 2026.

    PyPI Stats, Recent downloads for openai-agents (Nov 2026)

    Exact wording at the source
    "last_month":18956914
  4. #0212,641,760

    The crewai package on PyPI was downloaded 2,641,760 times in the month to November 12, 2026, about 1.7% of langchain's volume.

    PyPI Stats, Recent downloads for crewai (Nov 2026)

    Exact wording at the source
    "last_month":2641760
  5. #0221,975,932

    The llama-index package on PyPI was downloaded 1,975,932 times in the month to November 12, 2026; LlamaIndex also ships many separate integration packages not counted here.

    PyPI Stats, Recent downloads for llama-index (Nov 2026)

    Exact wording at the source
    "last_month":1975932
  6. #023423,864

    The autogen-agentchat package on PyPI was downloaded 423,864 times in the month to November 12, 2026, the lowest of the major frameworks tracked here.

    PyPI Stats, Recent downloads for autogen-agentchat (Nov 2026)

    Exact wording at the source
    "last_month":423864
  7. #02490M

    LangChain said in October 2025 that its langchain and langgraph libraries had a combined 90 million monthly downloads.

    LangChain, LangChain raises $125M to build the platform for agent engineering (Oct 2025)

    Exact wording at the source
    Today, langchain and langgraph have a combined 90M monthly downloads, and 35 percent of the Fortune 500 use our services.
  8. #025350M+

    LangChain's website reported more than 350 million monthly open source downloads across its packages in November 2026.

    LangChain, LangChain homepage (Nov 2026)

    Exact wording at the source
    350M+ Monthly open source downloads 7K+ Active LangSmith customers
  9. #02625M+

    LlamaIndex reported more than 25 million package downloads a month in November 2026, up from more than 3 million a month in March 2025.

    LlamaIndex, LlamaIndex homepage (Nov 2026)

    Exact wording at the source
    25M+ package downloads a month 300k+ LlamaParse users
  10. #02715 million

    Pydantic AI had been downloaded 15 million times in the nine months before its version 1 release in 2025.

    Pydantic, Pydantic AI V1 (Sep 2025)

    Exact wording at the source
    and a stunning 15 million downloads
  11. #028150K

    AWS's Strands Agents SDK passed 150,000 PyPI downloads and 2,000 GitHub stars within about two months of its May 2025 preview launch.

    Amazon Web Services, Introducing Strands Agents 1.0: Production-Ready Multi-Agent Orchestration Made Simple (Jul 2025)

    Exact wording at the source
    Since launching as a preview in May 2025, we've seen over 2,000 stars on GitHub and over 150K downloads on PyPI.

CH 04npm downloads for TypeScript agent frameworks

7 figures
  1. #029139,881,777

    Vercel's ai package, the AI SDK for TypeScript, was downloaded 139,881,777 times from npm in October 2026.

    npm, npm downloads for ai, October 2026 (Nov 2026)

    Exact wording at the source
    "downloads":139881777
  2. #03028,254,589

    The @langchain/core package was downloaded 28,254,589 times from npm in October 2026.

    npm, npm downloads for @langchain/core, October 2026 (Nov 2026)

    Exact wording at the source
    "downloads":28254589
  3. #03114,039,170

    The langchain package was downloaded 14,039,170 times from npm in October 2026.

    npm, npm downloads for langchain, October 2026 (Nov 2026)

    Exact wording at the source
    "downloads":14039170
  4. #0329,057,913

    OpenAI's @openai/agents package was downloaded 9,057,913 times from npm in October 2026.

    npm, npm downloads for @openai/agents, October 2026 (Nov 2026)

    Exact wording at the source
    "downloads":9057913
  5. #033619,037

    The llamaindex TypeScript package was downloaded 619,037 times from npm in October 2026, a small fraction of LlamaIndex's Python volume.

    npm, npm downloads for llamaindex, October 2026 (Nov 2026)

    Exact wording at the source
    "downloads":619037
  6. #03420 million

    Vercel reported more than 20 million monthly downloads for its AI SDK in December 2025.

    Vercel, AI SDK 6 (Dec 2025)

    Exact wording at the source
    With over 20 million monthly downloads and adoption by teams ranging from startups to Fortune 500 companies, the AI SDK is the leading TypeScript toolkit for building AI applications.
  7. #035220k

    Mastra, a TypeScript agent framework, reported more than 220,000 weekly npm downloads in November 2025, one year after launch.

    Mastra, Announcing Mastra 1.0 beta (Nov 2025)

    Exact wording at the source
    the leading Typescript agent framework with over 220k weekly npm downloads

CH 05Enterprise adoption and company scale

8 figures
  1. #03635 percent

    35% of Fortune 500 companies used LangChain's services as of October 2025, according to the company.

    LangChain, LangChain raises $125M to build the platform for agent engineering (Oct 2025)

    Exact wording at the source
    Today, langchain and langgraph have a combined 90M monthly downloads, and 35 percent of the Fortune 500 use our services.
  2. #0377K+

    LangChain reported more than 7,000 active customers for its commercial LangSmith platform in November 2026, including 5 of the Fortune 10.

    LangChain, LangChain homepage (Nov 2026)

    Exact wording at the source
    7K+ Active LangSmith customers 5 Of the Fortune 10 are LangSmith customers
  3. #03865%

    CrewAI said in November 2026 that 65% of the Fortune 500 used its platform, up from nearly half in October 2024.

    CrewAI, CrewAI homepage (Nov 2026)

    Exact wording at the source
    Used by 65% of the Fortune 500.
  4. #03910 million

    CrewAI's open source framework executed more than 10 million agents per month in October 2024, the company said when it announced funding.

    CrewAI, CrewAI Launches Multi-Agentic Platform to Deliver on the Promise of Generative AI for Enterprise (Oct 2024)

    Exact wording at the source
    CrewAI's open-source offering executes 10 million+ agents per month and is already used by nearly half of the Fortune 500.
  5. #040$18 million

    CrewAI raised $18 million in total funding by October 2024, including a Series A led by Insight Partners.

    CrewAI, CrewAI Launches Multi-Agentic Platform to Deliver on the Promise of Generative AI for Enterprise (Oct 2024)

    Exact wording at the source
    today announced $18 million in total funding that included an inception round led by boldstart ventures and a Series A led by global software investor Insight Partners.
  6. #041450 million

    CrewAI was running around 450 million agents a month in late 2025, its CEO told investor Insight Partners, about 45 times its October 2024 volume.

    Insight Partners, How CrewAI is orchestrating the next generation of AI Agents (Dec 2025)

    Exact wording at the source
    Nowadays, we're running around 450 million Agents a month
  7. #0421.4B

    CrewAI had powered 1.4 billion agentic automations by late 2025, according to Insight Partners.

    Insight Partners, How CrewAI is orchestrating the next generation of AI Agents (Dec 2025)

    Exact wording at the source
    now powering 1.4B agentic automations across the world's largest enterprises, including PwC, IBM, Capgemini, and NVIDIA.
  8. #043300k+

    LlamaIndex reported more than 300,000 users of its LlamaParse document parsing service in November 2026.

    LlamaIndex, LlamaIndex homepage (Nov 2026)

    Exact wording at the source
    25M+ package downloads a month 300k+ LlamaParse users

CH 06Agents in production and the tooling around them

6 figures
  1. #04457.3%

    57.3% of 1,340 professionals surveyed by LangChain in late 2025 had agents running in production, and another 30.4% were developing agents with plans to deploy.

    LangChain, State of Agent Engineering (Dec 2025)

    Exact wording at the source
    More than half of respondents surveyed (57.3%) now have agents running in production environments, with another 30.4% actively developing agents with concrete plans to deploy them.
  2. #04589%

    89% of organizations in LangChain's late 2025 survey had some form of observability for their agents, and 62% had detailed step-level tracing.

    LangChain, State of Agent Engineering (Dec 2025)

    Exact wording at the source
    89% of organizations have implemented some form of observability for their agents, and 62% have detailed tracing that allows them to inspect individual agent steps and tool calls.
  3. #04652.4%

    52.4% of organizations in LangChain's late 2025 survey ran offline evaluations of their agents on test sets.

    LangChain, State of Agent Engineering (Dec 2025)

    Exact wording at the source
    Just over half of organization (52.4%) report running offline evaluations on test sets
  4. #04726.5%

    Customer service was the most common primary agent use case in LangChain's late 2025 survey at 26.5%, followed by research and data analysis at 24.4%.

    LangChain, State of Agent Engineering (Dec 2025)

    Exact wording at the source
    Customer service emerged as the most common agent use case (26.5%), with research & data analysis close behind (24.4%).
  5. #04852%

    52% of developers in Stack Overflow's 2025 survey either did not use AI agents or used only simpler AI tools.

    Stack Overflow, 2025 Stack Overflow Developer Survey: AI (Jul 2025)

    Exact wording at the source
    A majority of developers (52%) either don't use agents or stick to simpler AI tools, and a significant portion (38%) have no plans to adopt them.
  6. #04917%

    17% of agent builders in Stack Overflow's 2026 survey used LangChain's LangSmith for observability, prompts or evals, second to Sentry at 26%.

    Stack Overflow, Stack Overflow Developer Survey 2026: AI agent observability and security (Jul 2026)

    Exact wording at the source
    Sentry leads at 26%, ahead of LangSmith at 17%.

Chart

Agent observability and eval tools used in the past year, 2026

Share of 2,277 Stack Overflow 2026 survey respondents who answered the agent observability question; multiple answers allowed.

  • Sentry26%
  • LangSmith17%
  • MLflow15%
  • Langfuse15%
  • Datadog LLM12%

Source: Stack Overflow, Stack Overflow Developer Survey 2026: AI agent observability and security

CH 07Protocols and interoperability

4 figures
  1. #05010,000

    There were more than 10,000 active public Model Context Protocol servers in December 2025, according to Anthropic, which created the protocol.

    Anthropic, Donating the Model Context Protocol and establishing the Agentic AI Foundation (Dec 2025)

    Exact wording at the source
    There are now more than 10,000 active public MCP servers, covering everything from developer tools to Fortune 500 deployments
  2. #05197M+

    Official MCP SDKs had more than 97 million monthly downloads across Python and TypeScript in December 2025.

    Anthropic, Donating the Model Context Protocol and establishing the Agentic AI Foundation (Dec 2025)

    Exact wording at the source
    with 97M+ monthly SDK downloads across Python and TypeScript
  3. #052407%

    The official MCP Registry grew 407% from its initial batch of servers within about two months of its September 2025 launch, to close to 2,000 entries.

    Model Context Protocol, One Year of MCP: November 2025 Spec Release (Nov 2025)

    Exact wording at the source
    That's a 407% growth from the initial batch of servers we onboarded that same month.
  4. #05337k

    The Model Context Protocol repository reached 37,000 GitHub stars in its first eight months, according to GitHub's 2025 Octoverse report.

    GitHub, Octoverse: A new developer joins GitHub every second as AI leads TypeScript to #1 (Oct 2025)

    Exact wording at the source
    Model Context Protocol (MCP) hit 37k stars in just eight months

Asked and answered

Questions readers ask

QWhat is the most popular AI agent framework?
ALangChain. 30% of agent builders in Stack Overflow's 2026 survey had used it, its main repository has 149.5k GitHub stars, and the langchain package was downloaded 157.2 million times on PyPI in the month to November 12, 2026. The OpenAI Agents SDK was second in the Stack Overflow survey at 24%.
QHow many GitHub stars do LangChain, CrewAI, AutoGen and LlamaIndex have?
AIn November 2026, LangChain had 149.5k stars, AutoGen 61.8k, CrewAI 60.9k and LlamaIndex 52.9k, according to their GitHub repository pages. LangGraph had 44.9k.
QHow many people download CrewAI?
AThe crewai package was downloaded 2,641,760 times on PyPI in the month to November 12, 2026, per PyPI Stats. CrewAI says its platform runs around 450 million agents a month and is used by 65% of the Fortune 500, though those are company figures.
QIs Microsoft AutoGen still maintained?
AMicrosoft has put AutoGen in maintenance mode and directs new feature work to Microsoft Agent Framework, which combines AutoGen with Semantic Kernel. AutoGen's share among agent builders in Stack Overflow's survey fell from 12% in 2025 to 4.5% in 2026, and its autogen-agentchat package drew 423,864 PyPI downloads in a month.
QHow many companies use AI agent frameworks in production?
A57.3% of 1,340 professionals in LangChain's late 2025 survey had agents in production, rising to 67% at organizations with more than 10,000 employees. Broader samples are lower: Stack Overflow's 2025 survey found 52% of developers did not use agents or used only simpler AI tools.

Where these numbers come from

We selected figures from primary publishers: developer surveys (Stack Overflow, JetBrains, IBM, LangChain), GitHub repository pages and the Octoverse report, the PyPI Stats and npm registry download APIs, and statements from the framework makers themselves. GitHub stars and PyPI monthly downloads were read on November 13, 2026 and change daily; npm counts cover October 1 to 31, 2026. Downloads include automated installs from CI systems and mirrors and count single packages, so frameworks that split code across many packages, such as LlamaIndex and LangChain, are undercounted per package and overcounted in company totals. Enterprise adoption figures are company claims without published definitions. Survey samples differ in size and audience, and LangChain's survey reaches its own users. We store the sentence each number appears in, and a script loads the publisher's page again to make sure that sentence is still there. Our methodology has the details.

Input log

Sources

  1. 01Amazon Web Services. Introducing Strands Agents 1.0: Production-Ready Multi-Agent Orchestration Made Simple (Jul 2025).
  2. 02Anthropic. Donating the Model Context Protocol and establishing the Agentic AI Foundation (Dec 2025).
  3. 03CrewAI. CrewAI homepage (Nov 2026).
  4. 04CrewAI. CrewAI Launches Multi-Agentic Platform to Deliver on the Promise of Generative AI for Enterprise (Oct 2024).
  5. 05GitHub. langchain-ai/langchain repository on GitHub (Nov 2026).
  6. 06GitHub. microsoft/autogen repository on GitHub (Nov 2026).
  7. 07GitHub. crewAIInc/crewAI repository on GitHub (Nov 2026).
  8. 08GitHub. run-llama/llama_index repository on GitHub (Nov 2026).
  9. 09GitHub. langchain-ai/langgraph repository on GitHub (Nov 2026).
  10. 10GitHub. Octoverse: A new developer joins GitHub every second as AI leads TypeScript to #1 (Oct 2025).
  11. 11IBM. AI Agents in 2025: Expectations vs. Reality (Mar 2025).
  12. 12Insight Partners. How CrewAI is orchestrating the next generation of AI Agents (Dec 2025).
  13. 13JetBrains. The State of Developer Ecosystem 2025 (Oct 2025).
  14. 14LangChain. LangChain raises $125M to build the platform for agent engineering (Oct 2025).
  15. 15LangChain. LangChain homepage (Nov 2026).
  16. 16LangChain. State of Agent Engineering (Dec 2025).
  17. 17LlamaIndex. Announcing our Series A and LlamaCloud general availability (Mar 2025).
  18. 18LlamaIndex. LlamaIndex homepage (Nov 2026).
  19. 19Mastra. Announcing Mastra 1.0 beta (Nov 2025).
  20. 20Model Context Protocol. One Year of MCP: November 2025 Spec Release (Nov 2025).
  21. 21npm. npm downloads for ai, October 2026 (Nov 2026).
  22. 22npm. npm downloads for @langchain/core, October 2026 (Nov 2026).
  23. 23npm. npm downloads for langchain, October 2026 (Nov 2026).
  24. 24npm. npm downloads for @openai/agents, October 2026 (Nov 2026).
  25. 25npm. npm downloads for llamaindex, October 2026 (Nov 2026).
  26. 26Pydantic. Pydantic AI V1 (Sep 2025).
  27. 27PyPI Stats. Recent downloads for langchain (Nov 2026).
  28. 28PyPI Stats. Recent downloads for langgraph (Nov 2026).
  29. 29PyPI Stats. Recent downloads for openai-agents (Nov 2026).
  30. 30PyPI Stats. Recent downloads for crewai (Nov 2026).
  31. 31PyPI Stats. Recent downloads for llama-index (Nov 2026).
  32. 32PyPI Stats. Recent downloads for autogen-agentchat (Nov 2026).
  33. 33Stack Overflow. Stack Overflow Developer Survey 2026: AI agent orchestration tools (Jul 2026).
  34. 34Stack Overflow. 2025 Stack Overflow Developer Survey: AI (Jul 2025).
  35. 35Stack Overflow. Stack Overflow Developer Survey 2026: AI agent observability and security (Jul 2026).
  36. 36Vercel. AI SDK 6 (Dec 2025).

Cite this page

Link to this page or to a single numbered figure. If you reuse a number, name the organization that released it in the first place.

DBotMaker. (2026, November 13). AI Agent Framework Statistics 2026. https://dbotmaker.io/ai-agent-framework-statistics

Download CSV

Found a number that changed or a source that moved? Email editorial@dbotmaker.io. We plan to check this page again by January 12, 2027.

Decisions that use this data

More readouts