LOCAL MODELS

An independent project fine-tunes Qwen3.5 and Gemma 4 into 0.8B and 2B decision models that answer in 22-28 ms using Jev's request shape. The verified numbers, the run commands, and where the benchmark stops matching real work.

Qwen3.8-27B is a 27B dense Apache-2.0 model that scores 61.7 on SWE-bench Pro and 42.2 on DeepSWE 1.1 - ahead of Opus 4.6 Max on both - while running on consumer hardware. Benchmarks, hardware math, and an honest when-to-use-it guide.

DeepReinforce AI released Ornith-1.0, a family of open-source coding models claiming self-improvement. The HN thread reveals a mix of skepticism and genuine interest - here is what the model actually does and whether the hype holds up.

JetBrains released Mellum2 on June 2, 2026 - a 12B MoE model with only 2.5B active parameters per token. Here is how to run it locally, when to use it, and where it fits in your AI coding stack.

Forge hit the Hacker News front page with a strong claim: small local models can become much more useful at tool-calling when the harness catches structural failures, retries intelligently, and controls context.

The trending Free Claude Code repo is not just about avoiding API bills. It points at a bigger developer-tool pattern: model gateways for AI coding agents.

How KV caching speeds up LLM inference - the math, the code, the memory tradeoffs, and when it stops helping. Every dev running local models hits this wall.

Qwen3.6-27B keeps pulling developers back because it sits in the awkward, useful middle: strong enough for real local coding tasks, small enough for serious workstation testing, and cheap enough to benchmark honestly.

Open-source AI code assistant for VS Code and JetBrains. Bring your own model - local or API. Tab autocomplete, chat, inline edit. Fully customizable.