Back to @PyTorch's profile

PyTorch engagement report

@PyTorch - 509K followers on X

Measured over 46 original posts from a 30-day window, last computed on August 28, 2026.

Engagement

Middle of its size range
Per follower
0.006%
of 509K followers
Per impression
0.206%
16K views on a typical post
Reach
3.10%
of its followers see a post
Typical post
32
interactions (median)
Saved
0.025%
4 bookmarks on a typical post
Posting rate
1.77/day
active 60% of days
Peak time
15:00 UTC
Thursday

A typical post picks up 32 interactions against 509K followers, an engagement rate of 0.006%. Measured over 46 original posts, its engagement rate beats 25% of 3,774 tracked accounts of a similar size, which puts it in the middle of its size range rather than at either end. Posts are seen about 16K times each, and 0.206% of those impressions turn into an interaction. That is about 3.10% of the follower count, which is the gap between an audience on paper and an audience in a timeline. Posting runs at about 1.8 posts a day over the last 30 days, with activity on roughly 60% of days. Most posts go out around 15:00 UTC, and Thursday is the busiest day of the week. Of the 46 posts sampled, 93% carry an image or video and 96% link out. The account's strongest tracked post pulled 198 interactions, about 6.2x its own typical post. Recurring topics include #pytorchcon, #ai, #cloudnativecon.

Measured over 46 original posts from a 30-day window, last computed on August 28, 2026. Recurring tags: #pytorchcon, #ai, #cloudnativecon.

Compared with accounts its own size

PyTorch's engagement rate beats 25% of the tracked X accounts closest to it in follower count (3,774 accounts, accounts of similar size (decile 7 of 10)). A percentile is spread evenly by construction, so 50 really is the middle of that group and 90 really is its top tenth.

On engagement per impression rather than per follower it beats 17% of the same group. When those two numbers disagree, the gap is about how far its posts travel rather than how people react to them.

Where this sits in the catalog

At 0.006%, PyTorch sits above the 10th percentile of the 36,521 accounts in this comparison. That places it in the bottom 25% band, which runs below 0.012%.

p100.002%
p250.012%
p50 (median)0.08%
p750.434%
p902.10%
p99160.7%
Engagement rate as a share of followers, across the 36,521 accounts we have scanned enough to measure. The axis is logarithmic, because the top and bottom of this population are about 107,166 times apart and a linear axis would flatten everything below the median into a single point.
Show the percentile table
Engagement rate percentiles
PercentileEngagement rate
10th percentile0.002%
25th percentile0.012%
50th percentile0.08%
75th percentile0.434%
90th percentile2.10%
99th percentile160.7%

This ruler is the whole measured catalog, not a size-matched group: it shows where the raw rate falls across every account we can measure, all of which are large. For a like-for-like comparison, read the size-band percentile above instead. See how the bands are built

Posting timing

This account posts most often around 15:00 UTC, and Thursday is its busiest day of the week. The bars below are the catalog-wide pattern, with this account's own busiest slot marked. They do not show how this account performs at each hour: we keep one aggregate per account, not one per hour, so that measurement does not exist in our data.

Engagement by hour posted, UTCTwenty-four bars, one per UTC hour. Each bar shows how posts published in that hour compare with their own authors' median engagement. Bars above the centre line ran higher than the median, bars below ran lower. A marker flags Busiest hour: 15:00 UTC.
0003060912151821
Above the authors' own mediansBelowScale: plus or minus 6%Busiest hour: 15:00 UTC
Show engagement by hour posted, utc as a table
Engagement by hour posted, UTC
Hour (UTC)Vs author medianPosts
00:00 UTC-1%50K
01:00 UTC-2%51K
02:00 UTC-3%50K
03:00 UTC-4%53K
04:00 UTC-6%43K
05:00 UTC-4%42K
06:00 UTC-4%48K
07:00 UTC-5%52K
08:00 UTC-4%60K
09:00 UTC-3%69K
10:00 UTC-2%72K
11:00 UTC-3%78K
12:00 UTC-2%86K
13:00 UTC-2%94K
14:00 UTC-4%97K
15:00 UTC-2%100K
16:00 UTC-3%97K
17:00 UTC-2%90K
18:00 UTC-1%84K
19:00 UTC-2%79K
20:00 UTC-1%74K
21:00 UTC-1%66K
22:00 UTC-2%57K
23:00 UTC-2%51K
Engagement by day of weekSeven bars, one per weekday, Sunday first. Each bar shows how posts published on that day compare with their own authors' median engagement. Bars above the centre line ran higher than the median, bars below ran lower. A marker flags Busiest day: Thursday.
SunMonTueWedThuFriSat
Above the authors' own mediansBelowScale: plus or minus 5%Busiest day: Thursday
Show engagement by day of week as a table
Engagement by day of week
DayVs author medianPosts
Sunday+4%230K
Monday0%286K
Tuesday-2%276K
Wednesday-1%251K
Thursday-1%244K
Friday-3%252K
Saturday+3%227K
See what moves engagement across the whole catalogWhat counts as a good engagement rate at this size

Best tweets

  • Aug 10, 20266.2x their median

    Today @AIatMeta introduced Muse Glimmer, an open-weight, 30-billion-parameter model distilled from Meta’s Muse Spark for on-device agentic workflows. Alongside, ExecuTorch is adding end-to-end support for running Muse Glimmer on @NVIDIA GPUs and Macs with @Apple silicon. Why ExecuTorch? ExecuTorch offers a smarter, native approach: ⚡ Build in PyTorch: Implement your model and decoding strategies right in PyTorch as you normally would ⚡ Seamless Export: Export directly to ExecuTorch when you're ready for deployment ⚡ Automated Optimization: The framework automatically handles backend-specific lowering (Triton on CUDA, MLX-native, or Metal on Apple silicon) and uses ahead-of-time compilation to optimize the full execution path end-to-end Learn more: https://t.co/ABzlzreEJz

    167199320K viewsView on X
  • Aug 13, 20264.5x their median

    AMD has been upstreaming optimizations for improved FP8 training support in PyTorch/TorchTitan and PyTorch/TorchAO, making FP8 training work out of the box on AMD Instinct GPUs! We trace that effort across every layer of the stack: - Systematically fusing Triton quantization kernels to close the performance gap, collapsing multi-kernel chains into single launches - Accelerating major kernel (grouped GEMM, attention, etc..) performance in PyTorch stack for FP8 training Read the blog: https://t.co/NysAnGQpTp @AMD

    117183520K viewsView on X
  • Aug 6, 20263.6x their median

    Every gain in pre-training efficiency means researchers can train larger models, run more experiments, and reach frontier capability faster on the same infrastructure. TorchTitan, PyTorch’s native training stack, leverages @NVIDIA contributions to continuously improve its performance on GB300 NVL72. These optimizations compound to deliver approximately 6x higher delivered performance on the same infrastructure, setting a record for pre-training DeepSeek-V3 671B at 1,648 TFLOPs per GPU. Read the full post: https://t.co/jYIP4tgrCE

    98115217K viewsView on X
  • Aug 25, 20262.9x their median

    And we have a winner. Congrats to @gatere_mark for creating the winning design in our 2026 PyTorch Foundation Flare Pin Community Design Contest! You'll be able to collect this flare pin at the Flare Party on Tuesday, October 20th during PyTorch Conference North America. Learn more & get your ticket to attend #PyTorchCon https://t.co/PW8IN4KiPU #PyTorchPin

    78132120K viewsView on X
  • Aug 22, 20262.8x their median

    "The PyTorch Conference is really unique in bringing together users, practitioners and developers across the AI spectrum in a way that really no other event does." Niles Burbank @AMD (PyTorch Conference Europe 2026) The open-source AI community is gathering in San Jose! Join us October 20–21 for PyTorch Conference North America 2026 for two days packed with technical sessions, hands-on workshops, and deep dives covering every layer of the AI stack. Register by September 4 to save $200 and secure your pass: https://t.co/m2JLV5jVNb #PyTorchCon

    671011018K viewsView on X
  • Aug 20, 20262.6x their median

    Leverage a PyTorch-native fine-tuning library with Day-0 Hugging Face checkpoint support for the recently released Alibaba open weights for Qwen3.8-2.4T-A95B (Qwen3.8-Max), its largest open-weight model. Developers can post-train the model for domain-specific use cases with configurable reasoning using PyTorch-native tools and NVIDIA NeMo AutoModel. Train directly on existing checkpoints without model conversion, with support for full SFT or memory-efficient LoRA fine-tuning. Read the full post: https://t.co/GFLEUcULkd

    6687117K viewsView on X
  • Aug 13, 20262.4x their median

    The PyTorch Certified Associate (PTCA) program offers early-stage practitioners a direct route to build credibility and demonstrate hands-on expertise in AI engineering. As deep learning adoption expands across industries, real-world skills in model design, training, and deployment are essential. PTCA provides a clear pathway to verify your foundational capabilities and stand out in the field. Learn more about the certification and how to enroll here 👉 https://t.co/BSnITGBEhi

    6674017K viewsView on X
  • Aug 27, 20262.3x their median

    We’re proud to announce the inaugural #PyTorchDayJapan, hosted by PyTorch Foundation, @huggingface, @IBM, and Mitsubishi Electric (@ME_JP_official), on December 10 in Tokyo. #PyTorchDayJapan will bring together PyTorch enthusiasts, machine learning engineers, AI researchers, and industry professionals for technical talks and discussions spanning training, inference, responsible AI, and more. PyTorch Day Japan is dedicated to open source AI and the impact of PyTorch Foundation projects including PyTorch, @vllm_project, @DeepSpeedAI, Ray, Helion, and Safetensors. 🔗 Learn more: https://t.co/w32fnU1tuu

    60130215K viewsView on X
  • Aug 21, 20262.3x their median

    PyTorch Conference North America 2026 lets contributors hear directly from users about the technical problems they're encountering and connect that feedback to ongoing work in PyTorch. In this clip from #PyTorchCon NA 2025, Tristan Rice @rice_fry (@Meta) describes how conversations with users helped him see that work in areas like fault tolerance and communication was paying off. #PyTorchCon North America returns October 20–21, 2026, in San Jose. Register by September 4 to save $200 and secure your pass: https://t.co/yIZoD2Avei

    6293017K viewsView on X
  • Aug 23, 20262.3x their median

    The keynote lineup for PyTorch Conference North America is out, offering a clear preview of where our open source ecosystem is heading. The program highlights four main directions: strengthening the core, expanding multi-hardware support, accelerating efficient training and inference, and building next-generation intelligent systems. Register now to see visionary keynotes from @Meta, @RedHat, @Google, @cohere, @AgenticAIFdn, @adaption_ai, @nvidia, @Qualcomm, @inferact, @coreauto, @CrusoeAI and @AWS. Learn more about our keynote speakers here: https://t.co/hPMKCE8DDP Register here: https://t.co/1z0jDhdUZm

    6363017K viewsView on X

Ranked by total interactions across everything we have tracked for this account, which is a longer history than the 30-day window the rates above use. The multiple compares each post to this account's own median.

Recurring topics

#pytorchcon#ai#cloudnativecon#kubecon#openinfrasummit

The most frequent hashtags in the sampled posts. They describe what this account writes about; they are not a performance signal, and the catalog-wide breakdown on the hub shows how little hashtag count moves.

Buy or sell X accounts - escrow-protected

PlayerSells is an escrow marketplace for X accounts. Every deal is protected, with no middleman risk.

Reading these numbers

A typical post picks up 32 interactions against 509K followers, an engagement rate of 0.006%. Measured over 46 original posts, its engagement rate beats 25% of 3,774 tracked accounts of a similar size, which puts it in the middle of its size range rather than at either end. Posts are seen about 16K times each, and 0.206% of those impressions turn into an interaction. That is about 3.10% of the follower count, which is the gap between an audience on paper and an audience in a timeline. Posting runs at about 1.8 posts a day over the last 30 days, with activity on roughly 60% of days. Most posts go out around 15:00 UTC, and Thursday is the busiest day of the week. Of the 46 posts sampled, 93% carry an image or video and 96% link out. The account's strongest tracked post pulled 198 interactions, about 6.2x its own typical post. Recurring topics include #pytorchcon, #ai, #cloudnativecon.

What is PyTorch's engagement rate on X?
PyTorch (@PyTorch) has an engagement rate of 0.006%, based on the median interactions across 46 original posts from the last 30 days against 508,564 followers. Replies, reposts and quote-posts of other people are excluded from that sample.
Is that a good engagement rate?
At 0.006%, PyTorch sits above the 10th percentile of the 36,521 accounts in this comparison. Those comparison accounts are all large ones, because our scanning cadence is weighted towards big accounts, so this is a ranking among peers of similar scale rather than a ranking across X.
Does @PyTorch have real engagement?
Its engagement rate beats 25% of the tracked X accounts closest to it in follower count (3,774 accounts), which puts it in the middle of its size range group. Ranking inside a size band matters because engagement rate falls as accounts grow, so a raw rate would mostly re-measure the follower count. It is a starting point for a look at follower quality, not a verdict on it.
When does @PyTorch post?
Most posts go out around 15:00 UTC, and Thursday is its busiest day, at roughly 1.77 posts per day across the measured window.

Keep going