cd/sources/dotnetperls-auto-discovered· home› sources› Dotnetperls (auto-discovered)
cat /sources/dotnetperls-auto-discovered.feed | wc -l → 79

Dotnetperls (auto-discovered)

articles 79 domain dotnetperls.com → page 1/4 feed RSS
07:00
2026-11-20
dotnetperls.com
large-language-models

Optimizing CSS with LLMs

A site owner cut 100-200 bytes from every page's inline CSS by pasting the stylesheet into the Muse Ember LLM and asking it to find size reductions, with no visual changes, at a total cost of 0.037 ce…

07:00
2026-11-18
dotnetperls.com
artificial-intelligence

Kev Decision Model in Llama

A blogger tested Kev, an open-weights decision model distributed as Kev-4B-Q4_K_M.gguf (a 3 GB file based on Qwen 3.5 4B) and served through Llama, which returns JSON answers with probabilities instea…

07:00
2026-11-15
dotnetperls.com
ai-tools

Installed Llama 1.0.0

A developer who had spent about 20 minutes compiling llama-cpp for each local LLM run over roughly a year installed the pre-built Llama 1.0.0 package, released a few days earlier, and got the inferenc…

07:00
2026-11-12
dotnetperls.com
large-language-models

Flint LLM in Pi on Rust Code

A hands-on test found that Recursive Superintelligence's Flint model, released last month with 2.5 billion active parameters out of 40 billion total, correctly removed all unsafe code from a Rust prog…

07:00
2026-11-06
dotnetperls.com
ai-tools

Quality of AI Code

A developer reports that switching from unstructured "vibe coding" to spec-driven development, in which an LLM generates code from a carefully maintained specification file, produced more reliable and…

07:00
2026-11-05
dotnetperls.com
artificial-intelligence

Vibe Coding Is Challenging

A developer reports that using Crouton Alpha to rewrite a hot Rust loop produced code that was slightly faster than previously hand-optimized code, but only after several failed attempts that showed p…

07:00
2026-11-02
dotnetperls.com
large-language-models

Beam 1.0 and Generating Functions

Reflection AI's open-weight Beam 1.0 model, with 501 billion parameters, generated 4 working Rust functions from specification files on the first attempt in a hands-on test, at under $0.002 per genera…

07:00
2026-10-26
dotnetperls.com
large-language-models

Kimi K3.1 Pro for Code Analysis

A hands-on test of Kimi K3.1 Pro on OpenRouter found the model took about 4 minutes and cost roughly $0.08 to analyze a small Rust function and return a simpler version, a cost the tester called somew…

07:00
2026-10-23
dotnetperls.com
large-language-models

Gemma 4 for Coding Help

A developer used the local model gemma-4-26B_q4_0-it.gguf in llama-cpp to fix long-standing HTML and JavaScript issues, adding lazily loaded images to a blog and a defer-loaded JavaScript file to a se…

07:00
2026-10-21
dotnetperls.com
large-language-models

Muse Wonder for Rust Code

A developer's test of Meta's Muse Wonder model on a Rust code regeneration task produced code that did not compile due to a lifetime issue in a closure, with the first response taking 54 seconds and c…

07:00
2026-10-19
dotnetperls.com
ai-tools

Refactored with MiniMax M3.1 Flash

MiniMax M3.1 Flash, the model formerly known as the stealth model Space Bunny, refactored a Rust codebase into a single match statement in 35 seconds, and the returned code worked correctly, according…

07:00
2026-10-17
dotnetperls.com
ai-tools

Using chain Function in Rust

A developer used Gemini 4 Argon to generate a Rust function that parses words from a Vector by calling iter on the Vector and chaining std::iter::once to append a space element, letting the loop itera…

07:00
2026-10-15
dotnetperls.com
large-language-models

Gemini 4 Argon for Blog Function

Google's Gemini 4 Argon generated working Rust code for a blog-post filename function on the first attempt, producing the output in 15.2 seconds using 1,178 tokens at a cost of $0.014, according to a …

07:00
2026-10-14
dotnetperls.com
ai-agents

Bonsai 1.7B for MCP Server

A hands-on comparison of two local LLMs for driving a custom MCP server found that Bonsai 1.7B, at 248.3 MB, was more reliable at tool calling than the smaller LFM 2.5 350M at 219.3 MB, despite runnin…

07:00
2026-10-11
dotnetperls.com
large-language-models

Claude Haiku Versus GPT-Luna

A blogger's comparison of Claude Haiku 5.5 and GPT-6-Luna on OpenRouter for Rust code generation found both models typically cost under $0.01 per generation (often as low as $0.005) and respond in 10-…

07:00
2026-10-10
dotnetperls.com
large-language-models

GPT-6-Luna and Specs

GPT-6-Luna condensed two Rust functions totaling roughly 80 lines into a single 30-line function in 21 seconds for $0.0011, according to a developer's account of a spec-driven code generation test. Th…

07:00
2026-10-09
dotnetperls.com
large-language-models

Idiomatic Rust Code and LLMs

A developer's hands-on comparison found that StepFun's Step 5 Preview and Mistral Large 4 generate working Rust code that is not idiomatic, relying on C-style `idx` local variables and byte accesses, …

07:00
2026-10-03
dotnetperls.com
large-language-models

Problems with Local LLMs

A developer's comparison found that generating complicated Rust functions on a local LLM took 30 to 60 minutes of 100% CPU and 100% GPU usage at an estimated electricity cost of $0.02 to $0.04, while …

07:00
2026-10-02
dotnetperls.com
ai-tools

Used Claude Sonnet 5.5 for Code Gen

A developer reported that Claude Sonnet 5.5 generated two Rust functions for about $0.06 after DeepSeek V4.1 Flash produced hard-to-read code and a slight performance regression. The developer said So…

07:00
2026-09-30
dotnetperls.com
artificial-intelligence

Rewrote Command Line Argument Handling

A developer replaced the Rust `clap` crate with LLM-generated argument-parsing code written by Gemini 3.8 Flash via OpenRouter, cutting the compiled executable size by 300 kB and argument parsing time…

page 1 / 4 next →