Goose with an uncensored model
To run Goose with an uncensored model, select its OpenAI provider and set OPENAI_HOST to an endpoint that serves an abliterated model.
.envShell
export GOOSE_PROVIDER=openai
export GOOSE_MODEL=kimi-k3.1-pro-uncensored
export OPENAI_HOST=https://api.silkcompute.com
export OPENAI_API_KEY=$SILK_API_KEY
goose session- OpenAI format
- OPENAI_HOST
- Anthropic format
- ANTHROPIC_HOST
- Tool calling
- MCP extensions
- Streaming
- Not documented
- Browser use
- Computer Controller
- Memory
- Memory extension
Labels link to Goose's own docs, read November 18, 2026. Compare all 11
Point the agent only at systems, files and accounts you own or are allowed to test.
Deploy it
3 steps, with Kimi K3.1 Pro.
- 01
Install the CLI. The desktop app works too.
Terminalcurl -fsSL https://github.com/aaif-goose/goose/releases/download/stable/download_cli.sh | bash - 02
Use the OpenAI provider with a custom host. OPENAI_BASE_PATH defaults to v1/chat/completions, which this endpoint serves.
See .env above - 03
Add extensions. Choose Toggle Extensions and turn on Memory and Computer Controller.
Terminalgoose configure
Which model
In this setup
Kimi K3.1 Pro
Long runs, image and video input
- Id
kimi-k3.1-pro-uncensored- Per 1M in / out
- $3.50 / $8.00
DeepSeek V4.1 Flash 1017
Value, image input
- Id
deepseek-v4.1-flash-1017-uncensored- Per 1M in / out
- $0.24 / $0.90
The config uses the Silk Compute base URL, which serves both models with a 1,048,576-token context.