llama server's default port changed from 8080 to 9931
retained text
Ref: [release](https://github.com/ggml-org/llama.cpp/releases/tag/v0.7.0), [PR](https://github.com/ggml-org/llama.cpp/pull/30159) Starting from `v0.7.0`, the default port for llama server is now `9931` (ggml). Those who did not explicitly set the port, expect breaking change upon upgrading. Those who use harnesses that have hardcoded `8080` port, good luck.
document fetched 2026-10-20 00:42 · canonical https://reddit.com/r/LocalLLaMA/comments/1x9tbg2/llama_servers_default_port_changed_from_8080_to