Local setup guide for OpenAI gpt-oss models with Codex CLI, Ollama, LM Studio, and MLX on Apple Silicon

Installation

Codex CLI

npm i -g @openai/codex
# or on macOS
brew install codex

Ollama — install from ollama.ai, then:

ollama pull gpt-oss:120b   # ~65 GB
ollama pull gpt-oss:20b    # ~13 GB (lighter alternative)

---

Running gpt-oss:120b

Interactive Ollama session

ollama run gpt-oss:120b

With Codex CLI (quick)

codex --oss --model gpt-oss:120b

With Codex CLI (configured profile)

Create/edit ~/.codex/config.toml:

[model_providers.ollama]
name = "Ollama"
base_url = "http://localhost:11434/v1"

[profiles.gpt-oss-120b-ollama]
model_provider = "ollama"
model = "gpt-oss:120b"

Then run:

codex --oss --profile gpt-oss-120b-ollama

---

Alternative Local Providers

LM Studio

[model_providers.lms]
name = "LM Studio"
base_url = "http://localhost:1234/v1"

[profiles.gpt-oss-120b-lms]
model_provider = "lms"
model = "gpt-oss:120b"

MLX (Apple Silicon)

pip install mlx-lm
mlx_lm.server --model SuperagenticAI/gpt-oss-20b-8bit-mlx --port 8888

[model_providers.mlx]
name = "MLX LM"
base_url = "http://localhost:8888/v1"

[profiles.gpt-oss-20b-8bit-mlx]
model_provider = "mlx"
model = "SuperagenticAI/gpt-oss-20b-8bit-mlx"

---

Remote Access (Tailscale/Network)

Bind Ollama to network

sudo mkdir -p /etc/systemd/system/ollama.service.d
sudo bash -c 'cat > /etc/systemd/system/ollama.service.d/override.conf <<EOF
[Service]
Environment="OLLAMA_HOST=0.0.0.0:11434"
EOF'
sudo systemctl daemon-reload
sudo systemctl restart ollama

Connect from another machine

export OLLAMA_HOST=<ip>:11434
ollama run gpt-oss:120b

# Or with Codex
CODEX_OSS_BASE_URL=http://<ip>:11434/v1 codex --oss --model gpt-oss:120b

---

Context Length

For larger projects, increase context window:

Provider	Method
Ollama	/set parameter num_ctx in session
LM Studio	lms load <model> --context-length <n>
MLX	Configure via server launch params

---

With LLM CLI

uv tool install llm
llm install llm-ollama
llm -m gpt-oss:120b "your prompt"

---

Hardware Requirements

• gpt-oss:120b — ~65 GB, needs substantial VRAM or RAM
• gpt-oss:20b — ~13 GB, more accessible
• gpt-oss-20b-8bit-mlx — quantized for Apple Silicon