Tags
Open-Weight
DeepSeek V4
A genuinely open, MIT-licensed model with frontier-tier coding results — download the weights, fine-tune them, or just use the cheap hosted …
GLM-5.2
Z.ai's MIT-licensed, 1M-token open-weight model — beats GPT-5.5 on FrontierSWE at a fraction of the cost, with full weights to self-host.
MiniMax M3
MiniMax's frontier open-weight model — 1M-token context, native multimodal input, and coding performance built for full-repository …
Qwen3-Coder-Next
The best practical local coding model — an Apache 2.0, mixture-of-experts model small enough to self-host on a single high-end machine.