Agent engineering, measured

Every figure here says where it came from

buildagentic tracks 46 open-source frameworks, language models and inference machines across 7 sources. Every value resolves to the source that reported it, the moment it was observed, and how it was measured. A figure that fails validation is withheld rather than shown — withheld means we refused to publish it, not that it is zero.
Datasets
3
Tracked entities
46
Sources
7
Last observation
just nowappend-only
AgnoStars42,358+5GitHub REST API · just nowLettaStars24,923+25GitHub REST API · just nowDeepEvalStars18,476+20GitHub REST API · just nowRagasStars15,862+6GitHub REST API · just nowPydantic AIStars20,220+20GitHub REST API · just nowsmolagentsStars29,530+24GitHub REST API · just nowMastraStars28,378+24GitHub REST API · just nowVercel AI SDKStars27,001+25GitHub REST API · just nowGuardrailsStars7,456+3GitHub REST API · just nowInstructorStars13,951+4GitHub REST API · just nowDSPyStars38,388+28GitHub REST API · just nowHaystackStars26,618+7GitHub REST API · just nowSemantic KernelStars28,608+1GitHub REST API · just nowAutoGenStars61,194+15GitHub REST API · just now

frameworks

Agent engineering frameworks

Open-source frameworks for building, orchestrating and evaluating LLM agents, tracked across GitHub, npm and PyPI.

Tracked
18
GitHub stars tracked
683,970
GitHub REST API
Last sync
just now · quarantined

models

Language and embedding models

Language, embedding and speech models tracked across Hugging Face Hub download activity and OpenRouter context windows and pricing.

Tracked
18
downloads, 30 days
69,145,892
Hugging Face Hub
Last sync
just now

hardware

Local inference hardware

GPUs, accelerators and SoCs for running models locally, with VRAM, memory bandwidth, power draw and list price.

Tracked
10
GB of VRAM catalogued
792
Last sync
just now

Biweekly

What changed, and what it means

New roles, the movements worth noticing across the datasets, and the occasional note on what broke. No filler.

Biweekly. Unsubscribe any time.