buildagentic

Tracker

Which agent frameworks are people actually using?

GitHub stars are a weak proxy for adoption and the only one every project publishes. Ranked lowest to highest, coloured by what each framework is for, so the shape of the field is visible before any single name is.
Every dot is a row in Agent engineering frameworks, with the source that reported it and the moment it was read.18 tracked →
GitHub stars
76 days of readings

LangChain leads on GitHub stars with 149k, 5.1x the median of 29k.

HighestLangChain · 149kMedianMastra · 29kLowestGuardrails · 8kTracked18

Dot height = GitHub stars, dot colour = what the framework is for. Left to right is rank, lowest first — the horizontal position carries no other meaning.

  • orchestration6
  • multi-agent3
  • evaluation2
  • rag2
  • structured-output2
  • memory1
  • optimisation1
  • validation1

memory, optimisation, validation share the neutral swatch — the palette carries six distinct colours. Point at a mark to name it.

037k75k112k149kGitHub starsLowestHighestGuardrails8kMastra29kLangChain149kGuardrails8kvalidationSynced from source, 2h agoInstructor14kstructured-outputSynced from source, 2h agoRagas16kevaluationSynced from source, 2h agoDeepEval19kevaluationSynced from source, 2h agoPydantic AI21kstructured-outputSynced from source, 2h agoLetta25kmemorySynced from source, 2h agoHaystack27kragSynced from source, 2h agoVercel AI SDK28korchestrationSynced from source, 2h agoSemantic Kernel29korchestrationSynced from source, 2h agoMastra29korchestrationSynced from source, 2h agosmolagents30kmulti-agentSynced from source, 2h agoDSPy39koptimisationSynced from source, 2h agoAgno43korchestrationSynced from source, 2h agoLangGraph45korchestrationSynced from source, 2h agoLlamaIndex53kragSynced from source, 2h agoCrewAI61kmulti-agentSynced from source, 2h agoAutoGen62kmulti-agentSynced from source, 2h agoLangChain149korchestrationSynced from source, 2h ago
Agent engineering frameworks →

History

What has moved since we started watching

76 distinct days of readings so far. Every observation is kept; none is overwritten.

  • LangChain

    149k +4k over 76d

    LangChain: 83 readings, 145k to 149k.

  • LangGraph

    45k +4k over 76d

    LangGraph: 83 readings, 40k to 45k.

  • CrewAI

    61k +3k over 76d

    CrewAI: 83 readings, 58k to 61k.

  • Mastra

    29k +2k over 76d

    Mastra: 83 readings, 27k to 29k.

  • Pydantic AI

    21k +2k over 76d

    Pydantic AI: 82 readings, 19k to 21k.

  • DeepEval

    19k +2k over 76d

    DeepEval: 82 readings, 18k to 19k.

  • Vercel AI SDK

    28k +1k over 76d

    Vercel AI SDK: 83 readings, 26k to 28k.

  • DSPy

    39k +1k over 76d

    DSPy: 83 readings, 38k to 39k.

  • smolagents

    30k +1k over 76d

    smolagents: 82 readings, 29k to 30k.

  • Agno

    43k +1k over 76d

    Agno: 82 readings, 42k to 43k.

  • AutoGen

    62k +1k over 76d

    AutoGen: 83 readings, 61k to 62k.

  • LlamaIndex

    53k +993 over 76d

    LlamaIndex: 83 readings, 52k to 53k.

  • Letta

    25k +967 over 76d

    Letta: 82 readings, 24k to 25k.

  • Ragas

    16k +804 over 76d

    Ragas: 81 readings, 15k to 16k.

  • Haystack

    27k +636 over 76d

    Haystack: 83 readings, 26k to 27k.

  • Instructor

    14k +326 over 76d

    Instructor: 82 readings, 14k to 14k.

  • Guardrails

    8k +272 over 76d

    Guardrails: 83 readings, 7k to 8k.

  • Semantic Kernel

    29k +254 over 76d

    Semantic Kernel: 83 readings, 28k to 29k.

More trackers

Other questions