# AI use cases for Open Jet (L-Forster/open-jet) Repo: [L-Forster/open-jet](https://github.com/L-Forster/open-jet) | Use cases: 3 | Page: https://dedupedai.com/use-cases/l-forster-open-jet/ OpenJet is a local AI coding agent and LLM inference engine that runs on your own GPU with no account or API key. `openjet setup` measures your hardware, picks and downloads a model and builds llama.cpp, and on RTX 30, 40 and 50 cards compiles its own CUDA megakernel for Qwen3.8-27B. It can serve 4 agents on one GPU, pair a frontier planner with a local worker (Slipstream), use Ollama or LM Studio, and ship a local model in your app via a Python SDK. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Run a coding agent fully offline on your own GPU: https://dedupedai.com/use-cases/l-forster-open-jet/run-a-local-coding-agent-on-your-gpu-with-openjet/ - Let Codex plan while a local model does the edits (Slipstream): https://dedupedai.com/use-cases/l-forster-open-jet/plan-with-codex-and-code-locally-with-openjet-slipstream/ - Use your existing Ollama or LM Studio models in OpenJet: https://dedupedai.com/use-cases/l-forster-open-jet/use-ollama-or-lm-studio-models-in-openjet/ ## How to run a coding agent fully offline on your own GPU with Open Jet Repo: [L-Forster/open-jet](https://github.com/L-Forster/open-jet) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Install OpenJet with pipx, let setup pick a model for your hardware, and start coding offline. **Who it is for:** A developer who wants a coding agent with no API bills and no code leaving the machine. **The problem:** Cloud coding agents cost per token and send your repo to a provider. **The outcome:** A local coding agent running a model sized to your VRAM, working offline after the download. ### Steps 1. Run `pipx install open-jet` 2. Run `openjet setup` and accept the recommended model (or `openjet setup --yes`) 3. On an RTX 3090, 4090 or 5090, choose Qwen3.8 27B Q4_K_M MTP so it builds the megakernel 4. Start the agent with `openjet` in your repo 5. If setup prints Skipping the Qwen3.8-27B megakernel, read the reason; it falls back to llama-server ### Prompt to paste into your agent ```text Install L-Forster/open-jet with pipx, run openjet setup for this machine's GPU, then use openjet in to . Tell me which model and engine it chose and why. ``` ### Caveat Setup downloads about 17 GB for Qwen3.8-27B and compiles llama.cpp; budget the disk space and time. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/l-forster-open-jet/run-a-local-coding-agent-on-your-gpu-with-openjet/ ## How to let Codex plan while a local model does the edits (Slipstream) with Open Jet Repo: [L-Forster/open-jet](https://github.com/L-Forster/open-jet) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Connect your ChatGPT Codex subscription or OpenRouter as the planner and let OpenJet's local model do the work. **Who it is for:** A developer who wants frontier-model planning but hits plan limits on long coding sessions. **The problem:** Frontier models burn through quotas when they read the whole repo and run every test themselves. **The outcome:** A frontier model that plans and reviews while the local model edits files and runs tests, using about 20% of the tokens. ### Steps 1. Install and set up OpenJet with `pipx install open-jet` and `openjet setup` 2. Inside OpenJet run /connect openai-codex (or /connect openrouter) 3. Switch with /mode to Slipstream 4. Give it a task; the local model reads the repo, edits files and runs tests 5. Review the frontier model's review of the local model's conclusions ### Prompt to paste into your agent ```text Set up L-Forster/open-jet in Slipstream mode with my Codex subscription as planner and the local model as worker, then do in and report tokens used by the planner. ``` ### Caveat This card is written from the L-Forster/open-jet README, not from a test run by Deduped. Before you start, make sure you have Linux with an NVIDIA RTX 30, 40 or 50 GPU with 24 GB+ and CUDA 12+ for the megakernel. Check the repo for newer instructions before you rely on these steps. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/l-forster-open-jet/plan-with-codex-and-code-locally-with-openjet-slipstream/ ## How to use your existing Ollama or LM Studio models in OpenJet with Open Jet Repo: [L-Forster/open-jet](https://github.com/L-Forster/open-jet) | Difficulty: Beginner | Time: 15 minutes | Basis: From the README Install the cloud extra and add a model profile pointing at your own OpenAI-compatible server. **Who it is for:** Someone who already runs Ollama or LM Studio and wants OpenJet's agent on top. **The problem:** Downloading another copy of the same model for a new tool wastes disk and setup time. **The outcome:** OpenJet using a model served by Ollama, LM Studio or your own llama-server. ### Steps 1. Run `pipx install 'open-jet[cloud]'` 2. Run `openjet --status` to find config.yaml 3. Add a model_profiles entry with runtime litellm, provider openai-compatible and your base_url, for example http://127.0.0.1:11434/v1 for Ollama 4. Select it in OpenJet with /model and the profile name 5. Note that OpenJet's hardware tuning and CUDA engine do not apply to external servers ### Prompt to paste into your agent ```text Configure L-Forster/open-jet to use my Ollama model through a model_profiles entry in config.yaml, select it with /model, and run in . ``` ### Caveat This card is written from the L-Forster/open-jet README, not from a test run by Deduped. Before you start, make sure you have Linux with an NVIDIA RTX 30, 40 or 50 GPU with 24 GB+ and CUDA 12+ for the megakernel. Check the repo for newer instructions before you rely on these steps. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/l-forster-open-jet/use-ollama-or-lm-studio-models-in-openjet/