Models
DeepSeek
Agent Arena Leaderboard Net Improvement

@arenaX · Aug 4, 2026

Agent Arena Leaderboard Net Improvement

@arena's benchmark bar chart drew 257.6K views on X: 5.6x its median post and #230 of 1,162 AI chart posts we hold by views. Read Oct 13, 2026.

DeepSeek V4 Flash ranks 21st with 2.0% net improvement

XData Benchmark bar chart DeepSeek V4 All DeepSeek models Kimi K3 Claude Opus 4.7 Claude Opus 4.6 Kimi-K2.7 Code Gemini-3.1 Pro DeepSeek-V4 ProClaude Opus 5 (High)Claude Opus 5 (Max)

@arena: Agent Arena Leaderboard Net Improvement — Published by @arena on X, Aug 4, 2026. Source: x.com

61/ 100

ViralChart score

Wide but thin

More views than the median post but a lower engagement rate: many saw it, few acted. Common for big accounts.

#432 of 1,162 AI chart posts we hold · #76 of 239 @arena posts. Score = views 40%, likes 30%, engagement rate 15%, save rate 15%, each as a percentile of every AI chart post we hold.

Our read

It sits in the Wide but thin quadrant, with a score rank of #432 of 1,162 AI chart posts.

Its 258K views are 5.6x @arena's median post, ranking #39 of 239 @arena posts by views.

Engagement sits at 0.41%, the 19th percentile among AI chart posts and below the 0.73% median.

The post ranked #1 of 4 @arena posts that day, ahead of the other 3.

5.6x@arena's median post

257.6K views, where @arena's typical post gets 46.2K.

5.8xthe median benchmark bar chart post

Across 215 benchmark bar chart posts we hold, the median gets 44.4K views.

4posts from @arena that day

Part of a run of 4 posts on Aug 4, 2026. This one was #1 of 4 by views that day.

bottom 19%engagement rate

0.41% of viewers liked, reposted, quoted or replied. The median post gets 0.73%.

Public X numbers from our read on Oct 13, 2026. Peers: 1,162 AI chart posts we hold and 239 from @arena.

Where it sits

Reach against resonance

Each dot is a post. Right is more views, up is a higher engagement rate. The lines are the medians, so the four corners are the four kinds of post.

Among 1,162 AI chart posts we hold

Score #432 of 1,162

Views	257.6K	#230	top 20%
Likes	905	#293	top 25%
Engagement rate	0.41%	#943	bottom 19%
Save rate	0.046%	#862	bottom 26%

Among 239 @arena posts

Score #76 of 239

Views	257.6K	#39	top 16%
Likes	905	#43	top 18%
Engagement rate	0.41%	#194	bottom 19%
Save rate	0.046%	#193	bottom 19%

Growth

Our reads cover 28 hours

Views257.6K

Age at our read70 days

Views per day of life3.7Kviews divided by days since it was posted

Gained between our reads+3views over 28 hours, about 3 a day now

Dots are our reads. The dotted line from the day it was posted to our first read is a gap, not data.

@arena's posts over time

Every post from this account we hold, by date and views. This one is marked; the line is the running median of the last 10.

News · Aug 4, 2026

@arena posts Agent Arena Leaderboard Net Improvement chart

The benchmark bar chart drew 258K views but an engagement rate of 0.41%, landing in the Wide but thin quadrant.

What happened

@arena posted a Benchmark bar chart on August 4, 2026, showing Agent Arena Leaderboard Net Improvement with 27 models. It came with 3 other posts from the account that day, including a Text-to-Image Arena Leaderboard at 137K views.

The numbers

It reached 258K views, ranking #230 of 1,162 AI chart posts by views and #39 of 239 @arena posts. Its score rank is #432 of 1,162 AI chart posts.

The lesson

A 27-model benchmark bar chart drew 258K views, 5.6x @arena's median, but an engagement rate of 0.41% against the 0.73% median among AI chart posts. The chart shows printed values from 12.1% to 9.2%.

What it means

For @arena, this post's 258K views are 5.6x its own median of 46.2K, yet its engagement rate of 0.41% is below the account's 0.68% median. The gap between views rank #39 of 239 and engagement rank #943 of 1,162 AI chart posts marks a wide but thin result.

Written by ViralChart from public numbers read on Oct 13, 2026. Every number is checked against our data before it is shown.

Updates: what @arena posted next

Data and API

30 points read off this chart

Estimated from the image 50%

Group	Value	Read
Claude Opus 5 (High)	12.1 %	printed
Claude Fable 5 (High)	11.7 %	printed
Kimi K3	11.5 %	printed
GPT-5.6 Sol (xHigh)	10.2 %	printed
Claude Opus 5 (Max)	10.1 %	printed
Claude Opus 4.8 (Thinking)	9.2 %	printed
GPT-5.5 (xHigh)	8.5 %	printed
Claude Opus 4.7 (Thinking)	8.2 %	printed
GPT-5.5 (High)	7.5 %	printed
Claude Sonnet 5 (High)	7.5 %	printed
Claude Opus 4.7	7.3 %	printed
GLM-5.2 (Max)	6.7 %	printed

Showing 12 of 30. All points, the redraw and the maker

Max plan · API

GET https://viral-chart.com/api/posts/5ede04d2da86.json
GET https://viral-chart.com/api/charts/5ede04d2da86.json
GET https://viral-chart.com/api/charts/5ede04d2da86.skill.md

Make one like this Open the post JSON Chart data JSON Skill file API docs

Static files in this prototype. The post JSON holds the public numbers, ranks, score, quadrant, our reads and, when checked, our copy.

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 5ede04d2da86.

See the Max plan API docs

	Path	Returns	Plan
GET	`/api/v1/charts/5ede04d2da86`	Title, publisher, date, form, point count; with a Max key, the data link	Open
GET	`/api/v1/charts/5ede04d2da86/data`	Every point with its provenance and confidence, plus the redraw spec. Add ?format=csv for CSV	Max
GET	`/api/v1/charts/5ede04d2da86/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

curlMax

curl -s https://viral-chart.com/api/v1/charts/5ede04d2da86/data \
  -H "Authorization: Bearer $VIRALCHART_KEY"

Claude CodeMax

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.