Models

Pinned Deepseek V4 then V4.1 Flash have been my workhorse models for months, both locally & on the APIs (but recently, GLM 5.3f, Mimo 2.6f, & Qwen 3.8f Next have been great alternatives for local hosting)
by Nick Antonaccio

Pinned Inexpensive LLM API providers for open source models
by Nick Antonaccio

Hy3 works locally on a single DGX Spark. It's a serious contender for best self-hosted model.
by Nick Antonaccio

Ternary models, which run on CPUs
by Nick Antonaccio

arcee-ai/trinity-large-thinking
by Nick Antonaccio

Laguna S 2.1 is a new self-hostable mid-size model.
by Nick Antonaccio

Don't forget about GLM 5.2
by Nick Antonaccio

Use MTP!
by Nick Antonaccio

Testing LLMs' depth of knowledge, all the way up to Anthropic Fable
by Nick Antonaccio

Qwen 3.6 35a3b & 27b are incredible for small GPUs, Gemma 4 26a4b & 31b are runners-up
by Nick Antonaccio

Step-3.7-Flash is surprisingly good
by Nick Antonaccio

Qwen 3.8 27b Issues
by Nick Antonaccio

GLM 5.3 may now be the best open source model
by Nick Antonaccio

Music Generation Models
by Nick Antonaccio

Mimo 2.6 Flash may become my default local LLM on a cluster of 2 DGX Sparks
by Nick Antonaccio

GPT 6 Astra is being rolled out - it beats Fable 5.1
by Nick Antonaccio

GLM 5.3 FLASH (OxAlpha) is a leader in local models. It's inexpensive + smart + fast on the API, originally an industry changing release, and still a killer buy.
by Nick Antonaccio

Deepseek 4.1 Flash is truly fantastic
by Nick Antonaccio

Comparison of self-hosted LLM Rubiks Cube demos
by Nick Antonaccio

Gemini 3.7 Flash is a real contender
by Nick Antonaccio