1. Chart data
  2. artificial-analysis
  3. AI Model Benchmarks
artificial-analysisBenchmark bar chart · 120 points · Grok 4

AI Model Benchmarks

The numbers behind a chart artificial-analysis published, redrawn natively. Check the original, study the data, then make your own.

Estimated from the image 51%Grok 4All xAI modelsAll chart data
Chart by artificial-analysis. Source: x.com

Simplified redraw

Drawn by us from the numbers below. Hover, tap or use the arrow keys.

A simplified redraw: it shows the numbers in our own look, not the look of the original. One accent marks what the chart is about. Values marked with ~ in the table are estimates.

The numbers

First 10 of 120 points are free
GroupSeriesValue (% / Tokens per second)SourceConf.
Claude 3.5 SonnetAA-Omniscence Hallucination Rate22Estimate0.97
Claude 3.5 SonnetIFBench Benchmark82.9Estimate0.97
Claude 3.5 SonnetOutput Speed492Estimate0.30
Claude 3.5 HaikuAA-Omniscence Hallucination Rate25Estimate0.97
Claude 3.5 HaikuIFBench Benchmark77.2Estimate0.90
Claude 3.5 HaikuOutput Speed304Estimate0.30
GPT-4oAA-Omniscence Hallucination Rate34Estimate0.30
GPT-4oIFBench Benchmark76.35Estimate0.90
GPT-4oOutput Speed268Estimate0.30
Gemini 1.5 ProAA-Omniscence Hallucination Rate40Estimate0.50

Max data, free with your email for now

See all 120 points and download the files

The full table, the JSON, a CSV and the agent skill file for this chart. These are Max in the plan; while this is a prototype your email unlocks them.

Unlocks all 35 giants. One email a week, leave any time. What you get · Log in

Gemini 1.5 ProIFBench Benchmark78Estimate0.50
Gemini 1.5 ProOutput Speed268Estimate0.50
Gemini 1.5 FlashAA-Omniscence Hallucination Rate48Estimate0.50
Gemini 1.5 FlashIFBench Benchmark77.2Estimate0.50
Gemini 1.5 FlashOutput Speed152Estimate0.50
GPT-4o miniAA-Omniscence Hallucination Rate51Estimate0.50
GPT-4o miniIFBench Benchmark77.1Estimate0.50
GPT-4o miniOutput Speed131Estimate0.50
Llama 3.1 405bAA-Omniscence Hallucination Rate56Estimate0.50
Llama 3.1 405bIFBench Benchmark75.4Estimate0.50
Llama 3.1 405bOutput Speed127Estimate0.50
Llama 3.1 70bAA-Omniscence Hallucination Rate61Estimate0.50
Llama 3.1 70bIFBench Benchmark75.4Estimate0.50
Llama 3.1 70bOutput Speed125Estimate0.50
Llama 3.1 8bAA-Omniscence Hallucination Rate64Estimate0.50
Llama 3.1 8bIFBench Benchmark73.9Estimate0.50
Llama 3.1 8bOutput Speed120Estimate0.50
GPT-4 TurboAA-Omniscence Hallucination Rate65Estimate0.50
GPT-4 TurboIFBench Benchmark72.3Estimate0.50
GPT-4 TurboOutput Speed91Estimate0.50
GPT-3.5 TurboAA-Omniscence Hallucination Rate72Estimate0.50
GPT-3.5 TurboIFBench Benchmark71.8Estimate0.50
GPT-3.5 TurboOutput Speed88Estimate0.50
Mistral Large 2AA-Omniscence Hallucination Rate80Estimate0.30
Mistral Large 2IFBench Benchmark70.95Estimate0.90
Mistral Large 2Output Speed78Estimate0.30
Mistral MediumAA-Omniscence Hallucination Rate83Estimate0.50
Mistral MediumIFBench Benchmark71.5Estimate0.50
Mistral MediumOutput Speed74Estimate0.50
Mistral SmallAA-Omniscence Hallucination Rate82Estimate0.50
Mistral SmallIFBench Benchmark70.2Estimate0.50
Mistral SmallOutput Speed73Estimate0.50
Qwen 2.5 72bAA-Omniscence Hallucination Rate87Estimate0.50
Qwen 2.5 72bIFBench Benchmark69Estimate0.50
Qwen 2.5 72bOutput Speed66Estimate0.50
Qwen 2.5 32bAA-Omniscence Hallucination Rate87Estimate0.50
Qwen 2.5 32bIFBench Benchmark60.7Estimate0.50
Qwen 2.5 32bOutput Speed58Estimate0.50
Qwen 2.5 14bAA-Omniscence Hallucination Rate87Estimate0.50
Qwen 2.5 14bIFBench Benchmark56.6Estimate0.50
Qwen 2.5 14bOutput Speed50Estimate0.50
Qwen 2.5 7bAA-Omniscence Hallucination Rate89Estimate0.50
Qwen 2.5 7bIFBench Benchmark53.7Estimate0.50
Qwen 2.5 7bOutput Speed49Estimate0.50
Qwen 2.5 3bAA-Omniscence Hallucination Rate89Estimate0.50
Qwen 2.5 3bIFBench Benchmark53.1Estimate0.50
Qwen 2.5 3bOutput Speed48Estimate0.50
Qwen 2.5 1.5bAA-Omniscence Hallucination Rate90Estimate0.50
Qwen 2.5 1.5bIFBench Benchmark62.7Estimate0.50
Qwen 2.5 1.5bOutput Speed44Estimate0.50
Qwen 2.5 0.5bAA-Omniscence Hallucination Rate91Estimate0.50
Qwen 2.5 0.5bIFBench Benchmark43Estimate0.50
Qwen 2.5 0.5bOutput Speed44Estimate0.50
Gemini 1.0 ProAA-Omniscence Hallucination Rate92Estimate0.50
Gemini 1.0 ProIFBench Benchmark42Estimate0.50
Gemini 1.0 ProOutput Speed44Estimate0.50
GLM-4.5AA-Omniscence Hallucination Rate34Estimate0.50
Gemini 2.5 FlashAA-Omniscence Hallucination Rate40Estimate0.50
MiniMax-M2AA-Omniscence Hallucination Rate45Estimate0.50
GPT-4.1AA-Omniscence Hallucination Rate55Estimate0.50
Gemini 2.5 ProAA-Omniscence Hallucination Rate60Estimate0.50
GPT-5AA-Omniscence Hallucination Rate61Estimate0.50
Claude 3.7 SonnetAA-Omniscence Hallucination Rate64Estimate0.50
Kimi K2AA-Omniscence Hallucination Rate65Estimate0.50
Grok 4AA-Omniscence Hallucination Rate72Estimate0.50
GPT-4.1 nanoAA-Omniscence Hallucination Rate80Estimate0.50
GPT-4.1 miniAA-Omniscence Hallucination Rate82Estimate0.50
Qwen3AA-Omniscence Hallucination Rate82Estimate0.50
Gemini 2.0 FlashAA-Omniscence Hallucination Rate87Estimate0.50
Claude 3.5 HaikuAA-Omniscence Hallucination Rate87Estimate0.50
Nova PremierAA-Omniscence Hallucination Rate89Estimate0.50
GPT-4.1AA-Omniscence Hallucination Rate89Estimate0.50
Claude 3.5 SonnetAA-Omniscence Hallucination Rate90Estimate0.50
Gemini 2.5 FlashAA-Omniscence Hallucination Rate91Estimate0.50
Gemini 2.5 ProAA-Omniscence Hallucination Rate92Estimate0.50
Nova PremierIFBench Benchmark: Results79Estimate0.50
Gemini 2.5 FlashIFBench Benchmark: Results78.8Estimate0.50
Gemini 2.5 ProIFBench Benchmark: Results78Estimate0.50
GPT-4.1IFBench Benchmark: Results77.2Estimate0.50
GPT-4.1 nanoIFBench Benchmark: Results77.1Estimate0.50
GPT-4.1 miniIFBench Benchmark: Results75.4Estimate0.50
Claude 3.7 SonnetIFBench Benchmark: Results72.3Estimate0.50
Grok 4IFBench Benchmark: Results71.8Estimate0.50
Qwen3IFBench Benchmark: Results71.6Estimate0.50
MiniMax-M2IFBench Benchmark: Results71.5Estimate0.50
Kimi K2IFBench Benchmark: Results69Estimate0.50
GLM-4.5IFBench Benchmark: Results60.7Estimate0.50
Gemini 2.0 FlashIFBench Benchmark: Results56.6Estimate0.50
Claude 3.5 HaikuIFBench Benchmark: Results53.7Estimate0.50
GPT-4.1IFBench Benchmark: Results53.7Estimate0.50
Claude 3.5 SonnetIFBench Benchmark: Results43Estimate0.50
Gemini 2.5 FlashIFBench Benchmark: Results42Estimate0.50
Nova PremierOutput Speed452Estimate0.50
Gemini 2.5 FlashOutput Speed304Estimate0.50
GPT-4.1Output Speed268Estimate0.50
GPT-4.1 miniOutput Speed268Estimate0.50
GPT-4.1 nanoOutput Speed266Estimate0.50
Grok 4Output Speed192Estimate0.50
Gemini 2.5 ProOutput Speed152Estimate0.50
MiniMax-M2Output Speed131Estimate0.50
Kimi K2Output Speed127Estimate0.50
GLM-4.5Output Speed120Estimate0.50
Qwen3Output Speed91Estimate0.50
Claude 3.7 SonnetOutput Speed78Estimate0.50
Gemini 2.0 FlashOutput Speed66Estimate0.50
GPT-4.1Output Speed58Estimate0.50
Claude 3.5 HaikuOutput Speed50Estimate0.50
Claude 3.5 SonnetOutput Speed49Estimate0.50
Gemini 2.5 FlashOutput Speed48Estimate0.50
Gemini 2.5 ProOutput Speed44Estimate0.50
Download this chart's data

Max subscribers get the same data by API. API docs

JSONAgent skill file

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 b5dce6c581d3.

See the Max planAPI docs
PathReturnsPlan
GET/api/v1/charts/b5dce6c581d3Title, publisher, date, form, point count; with a Max key, the data linkOpen
GET/api/v1/charts/b5dce6c581d3/dataEvery point with its provenance and confidence, plus the redraw spec. Add ?format=csv for CSVMax
GET/api/v1/charts/b5dce6c581d3/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/b5dce6c581d3/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.

Make one like this

This chart's shape, your numbers.

Live preview. Everything stays on your device.

Pick the form

A simplified form: our own look, not the look of the original. Your numbers start from this chart's data.

More data from xAI

All chart data
Estimated from the imageReplica 8.1Humanity's Last ExamxAI · Bar chart · 7 pointsEstimated from the imageCaseLaw BenchmarkxAI · Table chart · 6 pointsEstimated from the imageArtificial Analysis Intelligence Indexartificial-analysis · Bar chart · 17 pointsEstimated from the imageCursorBench 4.0 scorexAI · Line chart · 17 pointsEstimated from the imageGrok-4.20-Beta1 ranks #1 in Search Arenaarena · Table chart · 40 pointsEstimated from the imageReplica 5.3Artificial Analysis Intelligence Index v4.1.1artificial-analysis · Bar chart · 29 points

Questions people ask

How accurate are these numbers?

They are estimated from the image by two independent passes, with confidence 0.51 on a 0 to 1 scale. Printed numbers read well; values read off an axis can be a few percent off.

Can I reuse this data?

The numbers describe a chart that artificial-analysis published. The image, layout and styling belong to them. Use the numbers with a credit to the publisher, and make your picture from your own data.

How do I get the file?

The JSON, CSV and agent skill file are under the email gate on this page. In this prototype they are static files at /api/charts/b5dce6c581d3.json and /api/charts/b5dce6c581d3.skill.md.