Feat/local model providers - #1

Merged

ravituringworks merged 11 commits into

feat/local-model-providers

Aug 9, 2026

Commits on Aug 8, 2026

feat(ai): serve chat + expansion from ollama and llama-server

ravituringworks
and
claude
committed
a832f9c
feat(ai): add xai, cerebras, fireworks and sambanova recipes

ravituringworks
and
claude
committed
1e9159b
fix(models): tier defaults follow chat_model when no Anthropic key resolves

ravituringworks
and
claude
committed
974e4b4
fix(subagent): run non-Anthropic models instead of refusing them

ravituringworks
and
claude
committed
89366ee
fix(ux): stop telling users a hosted key is required when it is not

ravituringworks
and
claude
committed
b3c3b75
test(ai): pin the keyless-brain invariant

ravituringworks
and
claude
committed
ba7a9d1
docs: guide for running gbrain on local models

ravituringworks
and
claude
committed
a3427e8
fix(models): don't tell a free local provider that Anthropic is cheaper

ravituringworks
and
claude
committed
67f9da2
docs: state which vibecody providers were not added, and why

ravituringworks
and
claude
committed
9d61e12
fix(agent): correct the help text, and re-anchor the #2753 round-trip guard

ravituringworks
and
claude
committed
a45f252

Commits on Aug 9, 2026

feat(models): assign local model tiers from the discovered Ollama fleet

ravituringworks
and
claude
committed
f53ce70