Benchmarking

Explore recommended reads, summarized by AI

Machine Learning Evaluation

Writing about how models are tested, what benchmarks measure, and debates over their validity.

LLM Benchmark Critiques

Software Performance

Writing about measuring and comparing the speed and efficiency of code, frameworks, and protocols.

Programming Language Benchmarks

Popular this week

JavaScript in Plain English

Sajankumar Vijayan

The NestJS Node adapter won every run. By how much depends on how you measure.

3d ago

AI Advances

Jyrki Huhta

Building Molly: Wintermute Is Dead, Molly Is in Purgatory (and Suffering from Schizophrenia)

1d ago

Data Science Collective

Ankit Agrawal

Best Open Source OCR Model in 2026.

Sep 14

Alden Do Rosario

I Benchmarked 14 LLMs for LLM-Wiki and OKF. The Ranking Was the Least Interesting Result.

Aug 19

Saurabh Anil Agrawal

How do I get a Genie Agent above 90% accuracy? A benchmark-driven tuning log

Oct 16

Latest stories

John Ciliberti

Beyond the Baseline: Performance Engineering a High-Performance Rust Compression Engine

8h ago

Andre Burgstahler

Same Model, Different Harness, Different Score

14h ago

Full Expert Developer

5 things Nolan Lawson’s 60-app agent benchmark says to React developers

14h ago

AI Advances

Jyrki Huhta

Building Molly: Wintermute Is Dead, Molly Is in Purgatory (and Suffering from Schizophrenia)

1d ago

Jlod

Teaching an LLM to solve a Rubik’s cube: where it gets stuck, and how it could cheat

1d ago

See more stories

Benchmarking

Machine Learning Evaluation

LLM Benchmark Critiques

Software Performance

Programming Language Benchmarks

Popular this week

The NestJS Node adapter won every run. By how much depends on how you measure.

Building Molly: Wintermute Is Dead, Molly Is in Purgatory (and Suffering from Schizophrenia)

Best Open Source OCR Model in 2026.

I Benchmarked 14 LLMs for LLM-Wiki and OKF. The Ranking Was the Least Interesting Result.

How do I get a Genie Agent above 90% accuracy? A benchmark-driven tuning log

Latest stories

Beyond the Baseline: Performance Engineering a High-Performance Rust Compression Engine

Same Model, Different Harness, Different Score

5 things Nolan Lawson’s 60-app agent benchmark says to React developers

Building Molly: Wintermute Is Dead, Molly Is in Purgatory (and Suffering from Schizophrenia)

Teaching an LLM to solve a Rubik’s cube: where it gets stuck, and how it could cheat

Who to follow

Popular this week

The NestJS Node adapter won every run. By how much depends on how you measure.

Building Molly: Wintermute Is Dead, Molly Is in Purgatory (and Suffering from Schizophrenia)

Best Open Source OCR Model in 2026.

I Benchmarked 14 LLMs for LLM-Wiki and OKF. The Ranking Was the Least Interesting Result.

How do I get a Genie Agent above 90% accuracy? A benchmark-driven tuning log

Latest stories

Beyond the Baseline: Performance Engineering a High-Performance Rust Compression Engine

Same Model, Different Harness, Different Score

5 things Nolan Lawson’s 60-app agent benchmark says to React developers

Building Molly: Wintermute Is Dead, Molly Is in Purgatory (and Suffering from Schizophrenia)

Teaching an LLM to solve a Rubik’s cube: where it gets stuck, and how it could cheat

Who to follow