1. Models
  2. Z.ai
  3. LLM Performance Evaluation: Agentic, Reasoning, and Coding B
Z.aiX · Jul 28, 2025 · official

LLM Performance Evaluation: Agentic, Reasoning, and Coding Benchmarks

Z.ai's benchmark bar chart drew 1.3M views on X: 6.7x its median post and #88 of 1,162 AI chart posts we hold by views. Read Oct 13, 2026.

LLM benchmark performance across agentic, reasoning, and coding tasks

XDataBenchmark bar chartGLM-4All Z.ai modelso3Grok 4Gemini 2.5 ProKimi K2GPT-4.1GLM-4.5Claude 4 Opuso4-mini (high)
Published by Z.ai on X, Jul 28, 2025. Source: x.com
74/ 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.

#132 of 1,162 AI chart posts we hold · #12 of 34 Z.ai 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

The post sits in the Wide but thin quadrant, with a score rank of #132 of 1,162 AI chart posts.

Its 1.3M views sit 6.7x above Z.ai's 201K median, ranking #8 of 34 Z.ai posts.

At 8.14 views per follower, it lands in the 96th percentile among posts with known followers.

Its 0.30% engagement rate is 9th percentile among AI chart posts, against a 0.73% median.

30xthe median benchmark bar chart post

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

6.7xZ.ai's median post

1.3M views, where Z.ai's typical post gets 201.2K.

8.13views per follower

1.3M views from an account with 165.3K followers.

bottom 9%engagement rate

0.30% 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 34 from Z.ai.

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 #132 of 1,162

Views1.3M#88top 8%
Likes3K#101top 9%
Engagement rate0.30%#1,053bottom 9%
Save rate0.084%#539top 46%

Among 34 Z.ai posts

Score #12 of 34

Views1.3M#8top 22%
Likes3K#11top 31%
Engagement rate0.30%#30bottom 13%
Save rate0.084%#16top 46%

Growth

Our reads cover 26 hours

Views1.3M
Age at our read442 days
Views per day of life3Kviews divided by days since it was posted
Gained between our reads+3views over 26 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.

Z.ai'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 · Jul 28, 2025

Z.ai posts LLM benchmark bar chart with 13 models

The chart shows benchmark scores for 13 models, led by o3 at 65, and drew 1.3M views against a 0.30% engagement rate.

What happened

Z.ai published a benchmark bar chart on July 28, 2025, covering LLM performance across agentic, reasoning, and coding tasks. That day the account posted 3 times, including posts on GLM-4.5's agentic coding experience and AI model performance vs parameters.

The numbers

It reached 1.3M views, ranking #88 of 1,162 AI chart posts and #8 of 34 Z.ai posts. Its score rank is #132 of 1,162 AI chart posts.

The lesson

A 13-model benchmark bar chart drew 1.3M views but a 0.30% engagement rate, 9th percentile among AI chart posts. The chart shows o3 at 65 and Grok 4 at 63.6, yet the save rate was 0.084%.

What it means

For Z.ai, 1.3M views is 6.7x its 201K median, but the 0.30% engagement rate sits below its own 0.56% median. The gap between #88 of 1,162 AI chart posts by views and #1,053 of 1,162 by engagement marks this as wide but thin.

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

Updates: what Z.ai posted next

  1. AI Model Performance vs Parameters20.9K views
  2. GLM-4.5's Experience with Agentic Coding in Real-world Development38K views
  3. GLM-4.5 ARC Benchmark Performance153.3K views
  4. Open-source LLMs Benchmarks299.8K views
  5. Accuracy Comparison60.3K views
  6. CC-Bench: GLM-4.5's Experience with Agentic Coding in Real-world32.8K views

Data and API

14 points read off this chart

Estimated from the image 72%
SeriesGroupValueRead
o365printed
Grok 463.6printed
GLM-4.563.2printed
Claude 4 Opus60.9printed
o4-mini (high)60.4printed
GLM-4.5-Air59.8printed
Claude 4 Sonnet59.2printed
Gemini 2.5 Pro58printed
Qwen3-235B-A22B-Thinking55.9printed
DeepSeek-R1-052853.1printed
Kimi K248.7printed
GPT-4.146.3printed

Showing 12 of 14. All points, the redraw and the maker

Max plan · API

GET https://viral-chart.com/api/posts/b824c5c31485.json
GET https://viral-chart.com/api/charts/b824c5c31485.json
GET https://viral-chart.com/api/charts/b824c5c31485.skill.md
Make one like thisOpen the post JSONChart data JSONSkill fileAPI 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 b824c5c31485.

See the Max planAPI docs
PathReturnsPlan
GET/api/v1/charts/b824c5c31485Title, publisher, date, form, point count; with a Max key, the data linkOpen
GET/api/v1/charts/b824c5c31485/dataEvery point with its provenance and confidence, plus the redraw spec. Add ?format=csv for CSVMax
GET/api/v1/charts/b824c5c31485/render?view=1The draw-only view this page usesOpen
POST/mcpTool get_chart_data with this id, for Claude, Cursor and other agentsMax
curlMax
curl -s https://viral-chart.com/api/v1/charts/b824c5c31485/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.

More charts about GLM-4

GLM-4: LLM Benchmarks Performance Evaluation
OfficialX · Z.aiJan 19, 2026
LLM Benchmarks Performance Evaluation1.7M views · 5.3K likes
Data
GLM-4: LLM Performance on 8 Benchmarks
OfficialX · Z.aiFeb 11, 2026
LLM Performance on 8 Benchmarks1.5M views · 5.3K likes
Data
GLM-4: LLM Benchmarks Performance Evaluation
OfficialX · Z.aiDec 22, 2025
LLM Benchmarks Performance Evaluation1.4M views · 3.3K likes
Data
GLM-4: LLM Performance on 8 Benchmarks
OfficialX · Z.aiSep 30, 2025
LLM Performance on 8 Benchmarks907.8K views · 1.8K likes
Data
GLM-4: Open-source LLMs Benchmarks Comparison
OfficialX · Z.aiDec 8, 2025
Open-source LLMs Benchmarks Comparison364.4K views · 1.8K likes
Data
GLM-4: Artificial Analysis Intelligence Index v3.0
OfficialX · Z.aiDec 28, 2025
Artificial Analysis Intelligence Index v3.0338.6K views · 1.2K likes
Data

More from Z.ai

GLM-5: Agentic Coding Performance by Effort Level
OfficialX · Z.aiJun 16, 2026
Agentic Coding Performance by Effort Level7.7M views · 13.1K likes
Data
GLM-5: LLM performance across six benchmarks
OfficialX · Z.aiAug 26, 2026
LLM performance across six benchmarks7M views · 23.8K likes
Data
GLM-5: LLM benchmark performance across nine tasks
OfficialX · Z.aiAug 14, 2026
LLM benchmark performance across nine tasks5.9M views · 19.3K likes
Data
GLM-5.1: Coding Performance Evaluation
OfficialX · Z.aiApr 7, 2026
Coding Performance Evaluation4.4M views · 10.7K likes
Data
GLM-5: Coding Evaluation
OfficialX · Z.aiMar 27, 2026
Coding Evaluation1.3M views · 5.4K likes
Data
GLM-5.3: LLM Performance on Harvey LAB
OfficialX · Z.aiOct 12, 2026
LLM Performance on Harvey LAB533.2K views · 6.3K likes
Data

Questions people ask

Where does this chart come from?

Z.ai published it on Jul 28, 2025 on X (x.com). We keep it here, with the publisher and date credited as text, so you can study it without leaving.

Can I make a chart like this?

Use the format, not the picture. Open its data page and use Make one like this: it starts from this chart's numbers and takes yours.