This Week's GitHub Stars Ranking

📘 Tutorials 2026-11-16 8 min read

Hermes Agent gained 1,870 stars in a week, reaching 263,629 total stars—still first in both total and growth, but up 394 fewer stars than last week; Anthropic Quickstarts gained 1,651 stars in a week, surging to second in growth, with 80% of that growth concentrated in the first three days; across the full pool, 115 comparable projects added a net 23,163 stars, up 0.401%, 7.6% faster than last week. The names and order of the top ten stayed completely unchanged for the second consecutive week, and among the 116 entries, only 14 moved by one position.

📜 Table of Contents

The top ten from last week are kept exactly as they were, and this week continues the same: still ten projects, same order, not a single rank has moved. Hermes Agent gained 1,870 stars over seven days, reaching 263,629 stars total—still first on both counts, but its absolute weekly gain is 394 stars fewer than last week. The runner-up on the gainers list has changed hands: Anthropic Quickstarts added 1,651 stars in a week, more than quadrupling last week's 317. Across the full pool, 115 comparable projects gained a net 23,163 stars, a 0.401% increase—7.6% faster than last week's 0.374%. The pool's overall pace has warmed up slightly, yet the rankings have only grown more rigid.

Data Scope

Data is pulled from the official GitHub API, with a snapshot taken on November 15, 2026 (UTC), compared against a snapshot from seven days earlier using the same methodology. The tracking pool has 116 entries, all of which returned values in both weeks—no additions, no dropouts. Two of them, Embedchain and Mem0, point to the same repository (a redirect after a rename), so their star counts are fully duplicated; one is removed from the totals, leaving 115 actually comparable entries. Weekly gain equals current star count minus the star count seven days prior, with no smoothing and no adjustment for weekend effects.

After removing the duplicate, the pool grew from 5.7723 million to 5.7954 million, a net gain of 23,163 stars, or 0.401%. Last week, using the same method, it was 21,531 stars and 0.374%. The median single-project gain is 53 stars, up from 51 last week; the mean is 201 stars, up from 187. The median and mean differ by nearly four times, which shows gains are still concentrated in a handful of projects: 7 of the 115 entries gained over a thousand stars in a week, and those 7 account for 43% of the net gain.

This Week's TOP10

RankProjectStarsWeekly Gain
1Hermes Agent263,629+1,870
2n8n210,606+756
3Ollama187,092+710
4Awesome Prompts177,557+1,012
5Transformers169,020+354
6Dify161,213+526
7Open WebUI158,750+614
8Langflow156,682+235
9Claude Code155,792+792
10LangChain149,573+356

The threshold for the top ten rose from 149,217 to 149,573—only 356 stars higher in a week. The week before it rose 378 stars; the week before that, because the tenth spot switched from ComfyUI to LangChain, the threshold jumped 5,914 stars in one go. For two weeks running it hasn't been able to move, and the one stuck at the door is LangChain itself: it added only 356 stars this week, a 0.24% gain, the lowest in the top ten.

Hermes Agent: No Peaks, No Gaps in Seven Days

Over the seven days it added 264, 386, 286, 241, 255, 176, and 262 stars respectively, totaling 1,870—a daily average of 267. Last week was 2,264 stars, a daily average of 323, so this week is down 17.4%. The curve's high point was 386 stars on November 10, its low 176 on November 14—a 210-star spread between the two ends, well within normal day-to-day fluctuation, with no sign of any single day of concentrated exposure.

It accounts for 8.1% of the pool's net gain, down from 10.5% last week. The share is falling because the denominator grew, not because it slowed down: the pool gained 1,632 more stars this week, while it gained only 394 fewer—others simply grew faster. At 263,629 stars total, it leads second-place n8n by 53,023 stars; last week that gap was 51,909, so it widened by another 1,114 stars in a week. At the current rate differential, the distance between the top two will keep growing.

Anthropic Quickstarts: 80% of the Gain in Three Days, Back to Normal for the Last Four

It added 1,651 stars this week, reaching 21,045 total and 77th place overall, up one spot from last week. Its weekly gain of 8.51% ranks first among projects with over three thousand stars; second-place PR Agent is at 4.09%. Last week it added only 317 stars, 80% of which landed on the final day—leaving open the question of whether that jump could be sustained.

This week gives a partial answer. November 9 saw 532 stars, the 10th 479, the 11th 342—1,353 stars over three days, 82.0% of the whole week. The following four days were 61, 47, 106, and 84 stars, a daily average of 74.5—higher than its prior baseline, but not in the same league as the first three days. At just over twenty thousand stars total, a single popular post, course mention, or documentation reference can push it up—and just as easily let it fall back. Next week's comparison base becomes 21,045 stars, where 1,000 stars is only 4.7%; to judge whether it has held its ground, it's enough to watch whether the daily average stays above 70 stars.

Codex CLI Speeds Up 47%, llama.cpp Doubles and Closes In

Codex CLI added 1,588 stars this week, up from 1,078 last week—47.3% faster, the second-biggest acceleration among the top ten gainers, behind Lobe Chat's 53.8%. The daily line is front-loaded: 484 stars on November 9 and 455 on the 10th, the two days accounting for 59.1%, with the following five days ranging from 112 to 159. At 135,404 stars total it sits 12th, 14,169 stars behind tenth-place LangChain—at this week's rate differential, roughly eleven weeks away.

llama.cpp is right behind it. It added 1,439 stars this week, up from 681 last week—more than double. At 135,201 stars total it sits 13th, 203 stars behind Codex; last week the two were only 54 stars apart. Its daily line is the exact inverse of Codex's: only 78 stars on November 9, then 190 on the 11th, 322 on the 12th, peaking at 404 on the 13th, and falling back to 189 on the 14th. One ships in the first two days, the other in the middle three—and the seven-day totals end up only two hundred stars apart. Which of these two holds which position next month is anyone's guess.

ComfyUI Slips to Fifth in Gains, Lobe Chat Up Over 40% in a Week

ComfyUI added 1,229 stars this week, down from 1,312 last week—a 6.3% decline—dropping from third to fifth on the gainers list. At 147,022 stars total it sits 11th, 2,551 stars behind tenth place; last week that gap was 3,424, narrowing by 873 in a week. At this week's differential of 873 stars per week, the tenth spot could change hands within three weeks. This is the loosest position in the top ten; the spacing between the other nine is either widening or holding still.

Lobe Chat is the most conspicuous one around 22nd place. It added 1,149 stars, up from 747 last week—53.8% faster—for 85,572 total, still ranked 22nd. November 10's single-day 282 stars was the week's high point; the final day added only 47. Its gap to 21st-place NextChat shrank from 4,519 stars to 3,396, a narrowing of 1,123; NextChat added only 26 stars this week, so at that differential it should pass in about three weeks. The mid-table rankings will be livelier than the top ten in the weeks ahead.

Who's Falling

Seventeen repositories lost stars, for a combined decline of 75, with another 3 showing zero growth. The biggest decliners were ChatGLM down 15, GPT Engineer down 11, Chatbot UI down 10, LocalStack down 8, GPT4All and Mistral Inference down 6 each, OpenChat down 4, and the remaining ten down 1 to 2 each. The zero-growth ones are Pinecone, Phi Cookbook, and Noi. At these magnitudes on repositories with tens of thousands of stars, this is routine unfollowing and says nothing about users leaving.

Deceleration is more worth watching than negative growth. Last week's second-biggest gainer, PR Agent, added 638 stars this week—1,095 fewer, a 63.2% drop, the fastest decline in the pool. It didn't exceed 150 stars on any day of the week; its peak was 147 on November 10, and the kind of single-day 470-star exposure it saw last week didn't recur. At 16,236 stars total, it moved from 87th to 86th—only because 86th fell further. The model repos themselves continue to hug the floor: DeepSeek V3 added 75 stars, R1 added 27, Qwen 2.5 added 12, and Yi and InternLM each lost 1.

Why the Rankings Won't Move

Of the 116 entries, 102 held their rank, 14 moved by one spot—seven up, seven down—and the biggest change in the entire pool was a single position. This isn't at odds with a net gain of 23,163 stars: the upper half of the pool is already above a hundred thousand stars, and the gap between tenth and eleventh place is just 2,551 stars. For anyone behind to catch up, they'd need to out-gain by over a thousand stars a week for several weeks running.

The four entries from 11th to 14th—ComfyUI, Codex CLI, llama.cpp, and Browser Use—added a combined 5,022 stars this week, the hottest stretch in the pool, yet their totals remain between 120,000 and 150,000, each 2,551 to 14,169 stars behind tenth place. The gainers list best shows where attention is going; the rankings list only shows that the size gap hasn't been eaten away yet. The two tables need to be read separately.

How to Use This List

Star counts reflect attention only, with no direct bearing on quality or performance. Use totals to gauge ecosystem maturity and weekly gains to gauge where recent attention is flowing—look at both, but don't treat either as an acceptance criterion. For real selection decisions, release cadence, issue response speed, and running it once in your own scenario are more useful than star counts. This week's numbers draw three lines: the top ten positions haven't budged for two weeks running, while the pool's overall pace is 7.6% faster than last week; gains are concentrating in three specific categories—coding CLIs, inference frameworks, and chat interfaces—with three new faces on the gainers list (Anthropic Quickstarts, Browser Use, n8n in; PR Agent, Ollama, Open WebUI out); and the model repos themselves remain near zero growth. Two things to watch next week: whether Anthropic Quickstarts' daily average holds above 70 stars, and whether ComfyUI can catch tenth place within three weeks.

Same Time Next Week

The list updates every Monday, with a fixed data source and methodology for easy week-over-week comparison. If you'd like to keep this table around to read at your leisure, bookmark this page; subscribe to site updates and next week's gainers list and ranking changes will reach you first. If any project you're using has had an unusual rise or fall this week, drop a line in the comments and we'll pull it out for a closer look next time.

## Related Articles - [This Week's GitHub Stars Ranking: Codex Leads with 1,865 Stars Gained](/post/github-ai-agent-stars-weekly-20261102) - [This Week's GitHub Stars Ranking: Hermes Agent Tops with 2,264 Stars Gained in a Week, Entire Top Ten Unchanged](/post/github-ai-agent-stars-weekly-20261109) - [Running Ollama on Android: A Comparison of 4 Apps](/post/run-ollama-locally-on-android-2026)
Related Articles
2026-08-13
Whisper Local Transcription 2026: Open Source Speech-to-Text Without the Cloud
2026-08-11
Small Language Models 2026: When a 1-8B Model Beats the Giants
2026-08-05
Free AI Grammar Checker Software in 2026: LanguageTool (14k Stars) Self-Hosted vs Grammarly - Private Writing Assistant
2026-07-19
LLM Inference Optimization: Make Your Model Run Faster
2026-08-05
Run Llama Locally in 2026: Complete llama.cpp (123k Stars) Guide - From GGUF Download to Chat on Any Hardware
2026-08-07
AI Image Upscaler: Real-ESRGAN and Friends for 4K Restoration

Written by our editorial team; tools listed here are tested or verified against public sources. Links point to official sites or GitHub repos for reference only — no paid placements.

💬 Comments (0)

No comments yet. Be the first!

Login to comment