Claude Code
Anthropic · cloud API
Defaults apply when this runtime is active. Model catalog: Models.
Strong for refactors and multi-file edits.
| Auto-apply safe edits | |
| Run approved shell commands | |
| Verbose tool traces | |
Claude Code in the terminal, with inference routed through Vertex AI instead of the Anthropic API. Uses agents.json env entries and optional gcloud mounts.
|
Publisher model ID for your region and contract (mock list). |
|
|
Written to environment for the Claude agent in Mounts ( No |
|
| Auto-apply safe edits | |
| Run approved shell commands | |
| Verbose tool traces | |
Editor-native; fast iteration on open files.
| Auto-apply safe edits | |
| Run approved shell commands | |
| Verbose tool traces | |
Block’s extensible agent for the terminal — providers, extensions, and tool use. Configure API keys and backends under Models / provider settings as needed.
|
Provider id mirrors what Goose expects for the active profile (mock). |
|
| Auto-apply safe edits | |
| Run approved shell commands | |
| Verbose tool traces | |
Runs on your machine; no vendor inference API.
|
VRAM / RAM mirrors the Models page for this host. |
|
| Auto-apply safe edits | |
| Run approved shell commands | |
| Verbose tool traces | |