Model Performance Comparison
Meta's line chart drew 1.9K views on X: 0.1x its median post and #1,082 of 1,162 AI chart posts we hold by views. Read Oct 12, 2026.
To quantify how the models handle unfamiliar tasks, we compared test-time compute scaling on tasks trained on during post-training against other tasks.
Why it spread
This chart post from @AIatMeta did not spread: it reached 1,955 views but only 5 likes, 0 reposts, 0 quotes flat, and 1 save, against an account typical likes of 133, which is 0.0x normal, so the measured numbers show almost no engagement despite the launch timing and polished design.
What worked
- The post text frames a clear comparison: test-time compute scaling on tasks trained on during post-training against other tasks, which gives the line chart a specific question to answer.
- The chart type is a line chart showing model performance gains from in-distribution to out-of-distribution, a form that suits a two-point or continuous comparison.
- The design polish is medium and the self labelling is medium, so the visual is readable and the post is branded, which supports comprehension even if reach stayed below average.
What to borrow
We could use the same framing by taking one of our own comparisons, such as performance on familiar versus unfamiliar tasks, and plotting it as a line chart with medium polish and medium self labelling, while making sure the post text states the comparison plainly.
Careful
Do not copy the assumption that a branded, polished chart will spread on its own: this post had below average source reach, a quote rate in the bottom quarter, no source line, and only 5 likes, 0 reposts, 0 quotes, and 1 save from 1,955 views, so do not over-read the launch timing or design as a success.
Written by our analysis model from the measured numbers above, not by Meta.
Make it yours
Your numbers, the bar chart way.
Starts from the real numbers of "Out-of-Distribution Performance vs In-Distribution Performance" by Meta. Replace every number with yours before you post. Make one like this, step by step
Free pictures carry a small "Made with ViralChart" line. That line is how other people find us.
API · Max
Use this chart via API
On the Max plan you can pull this chart's numbers with the provenance of each one, as JSON or CSV, with its redraw spec, or call it from an agent over MCP. Chart id 085ff9acc300.
| Path | Returns | Plan | |
|---|---|---|---|
| GET | /api/v1/charts/085ff9acc300 | Title, publisher, date, form, point count; with a Max key, the data link | Open |
| GET | /api/v1/charts/085ff9acc300/data | Every point with its provenance and confidence, plus the redraw spec. Add ?format=csv for CSV | Max |
| GET | /api/v1/charts/085ff9acc300/render?view=1 | The draw-only view this page uses | Open |
| POST | /mcp | Tool get_chart_data with this id, for Claude, Cursor and other agents | Max |
curl -s https://viral-chart.com/api/v1/charts/085ff9acc300/data \
-H "Authorization: Bearer $VIRALCHART_KEY"claude mcp add --transport http viralchart https://viral-chart.com/mcp \
--header "Authorization: Bearer $VIRALCHART_KEY"60 requests a minute per key; over that you get 429 with Retry-After. Up to 5 keys per account, made on your account page and shown once. Without Max, the data route answers 402 with a link to the plans. The numbers describe a chart its publisher owns, and every file carries a license note.
More from Meta
All 11
Other giants using a line chart
All 295
Questions people ask
How far did Meta's "Model Performance Comparison" chart travel?
1,955 views, 5 likes, 0 reposts and 1 saves on X, read Oct 12, 2026. A normal post from @AIatMeta gets about 25.7K views.
What kind of chart is it?
A line chart. Model performance gains from in-distribution to out-of-distribution.
Can I use Meta's chart?
Use the format, not the picture. The chart, logo and design belong to Meta. ViralChart redraws the chart type with your numbers and words, and credits the original publisher.