ORNITH-K8S — LOCAL DEVOPS AGENT FOR KUBERNETES OPERATIONS

A custom model built on DeepReinforce Ornith-1.0-9B, tuned as an autonomous and automation agent. It speaks the Anthropic-compatible API, so it drives Claude Code, Codex and opencode fully locally

• your code never leaves your machine and cloud token costs drop to zero.

Ornith-1.0-9B scores 69.4% on SWE-Bench Verified - beating models 3-4x larger (Gemma 4-31B, Qwen 3.6-35B) in coding tasks, despite having only 9B parameters.

The model ships with a system prompt focused on real work in a terminal: use tools instead of guessing, write files instead of pasting code, ground every answer in real tool output, and stay terse. Thinking is suppressed (think:false), context set to 64K.

MODELS IN THE FAMILY

Model | Base | Context | Size | Purpose ornith-claude-coder:9b | Ornith-1.0-9B (GGUF Q4_K_M) | 64K | 5.6 GB | Coding agent, SWE-Bench 69.4%, 32GB+ Apple Silicon

WHAT IT'S FOR

• Driving Claude Code / Codex / opencode locally
• Agentic code writing and editing with native function calling / tool use
• Sysadmin / devops tasks in a real terminal (disk, network, scripts)
• Full privacy and offline operation

QUICK START

$ ollama run ornith-k8s

In Claude Code: $ ollama launch claude --model ornith-k8s

TESTS

## User

execute kubectl to get nodes on the current context

---

## Assistant (Build · Ornith K8S · 63.5s)

_Thinking:_

The user wants me to run a simple kubectl command to list nodes. Straightforward.

**Tool: bash**

**Input:**
```json
{
  "command": "kubectl get nodes"
}

Output:

NAME   STATUS   ROLES                  AGE    VERSION
k8s    Ready    control-plane,worker   473d   v1.34.6

BEHAVIOR TUNING

• No thinking. SYSTEM /nothink + think:false in API. Model acts, does not monologue.
• No hallucination. Reports only values from tool output
  • does not fabricate.
• Acts, never asks. Inspect / scan / check / measure -> runs the command.
• Terse, one language. No preamble, no recap, matches the user's language.

SAMPLING / CONTEXT

temperature 0.2, top_p 0.9, top_k 20, repeat_penalty 1.05, num_ctx 65536 Native context 262K - can be raised on stronger hardware

HOW IT WAS MADE

Duplicate from rafw007/ornith-1.0-9b-claude-coder with adaption for Devops and Kubernetes Operations.

LICENSE

MIT (inherited from the base DeepReinforce Ornith-1.0 model).

--- Card: 2026-08-20. Tests on Mac Mini and opencode