diff options
| author | Paul Buetow <paul@buetow.org> | 2026-05-01 15:10:20 +0300 |
|---|---|---|
| committer | Paul Buetow <paul@buetow.org> | 2026-05-01 15:10:20 +0300 |
| commit | e5cd3949c477308709f0052bbd0fae68a970bfa7 (patch) | |
| tree | c065f59f5976943874ba1a90b63c7decae1fe494 | |
| parent | c9f3180fda2c07e5a27b1a99e74c5e9ec83b4222 (diff) | |
hexai: switch default Ollama Cloud model to gemma4:31b-cloud
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
| -rw-r--r-- | hexai/config.toml | 9 |
1 files changed, 4 insertions, 5 deletions
diff --git a/hexai/config.toml b/hexai/config.toml index 388faf2..1a6dff1 100644 --- a/hexai/config.toml +++ b/hexai/config.toml @@ -16,14 +16,13 @@ slashcommand_dir = "~/.cursor/commands" name = "ollama" [ollama] -model = "kimi-k2.6" +model = "gemma4:31b-cloud" base_url = "https://ollama.com" temperature = 0.2 -# In-code auto-completion uses gemma4:31b (the dense larger Gemma 4 on Ollama -# Cloud). It's faster and tighter than kimi-k2.6 for short, latency-sensitive -# completions while everything else still defaults to kimi-k2.6 above. +# In-code auto-completion uses gemma4:31b-cloud (the dense Gemma 4 hosted on +# Ollama Cloud). Latency-sensitive completions use the same model as chat. [[models.completion]] provider = "ollama" -model = "gemma4:31b" +model = "gemma4:31b-cloud" temperature = 0.2 |