Skip to content
inferamdocs
Website ↗

Coding agents

omp

Add Inferam to omp (oh-my-pi) as a provider, next to the ones you already use, with one command.

omp (oh-my-pi) talks to Inferam over OpenAI Chat Completions.

Terminal window
inferam setup omp

Inferam is now one of omp’s providers, every time you run omp. Pick DeepSeek-V4.1-Flash-1017 under inferam with /model, or run omp --model inferam/DeepSeek-V4.1-Flash-1017, and switch to any other provider whenever you like, in the same session. Your default model and your other providers stay as they were.

If omp isn’t installed, setup offers to install it first, with omp’s own installer (inferam --yes setup omp installs it without asking). The first time it runs, setup also signs you in if this computer isn’t signed in yet.

Terminal window
inferam omp

This starts omp on Inferam’s model, unless you pass --model or --provider, with your other providers listed as usual. The first time, it runs inferam setup omp for you, so you don’t need both. Arguments pass through: inferam omp -p "summarise the diff", inferam omp --continue, inferam omp --resume <id>, inferam omp --profile work.

Sessions start at Inferam’s default reasoning effort, high. Change it in omp as usual; every level from minimal to max is available.

One file in omp’s agent folder, ~/.omp/agent (or $PI_CODING_AGENT_DIR; with a profile, from --profile or $OMP_PROFILE, ~/.omp/profiles/<name>/agent):

models.yml (or your models.yaml, if that’s the one you have) gets an inferam entry under providers. Your other providers, settings and comments are kept:

providers:
inferam:
baseUrl: https://api.inferam.ai/v1
api: openai-completions
apiKey: ifr_...
models:
- id: DeepSeek-V4.1-Flash-1017
name: DeepSeek-V4.1-Flash-1017
reasoning: true
thinking:
mode: effort
efforts: [minimal, low, medium, high, xhigh, max]
defaultLevel: high
input: [text]
contextWindow: 262144
maxTokens: 32768
cost: {input: 0, output: 0, cacheRead: 0, cacheWrite: 0}

apiKey is omp’s own Inferam key, named “omp on <hostname>” on inferam.ai/keys. omp takes a provider’s key in its entry in this file, so that’s where it goes, and the file is readable only by you, as omp writes it itself. Don’t share or commit this file.

If you only have a models.json from pi, setup starts models.yml from it, as omp would on its next start, so your providers carry over. models.json itself isn’t changed.

Running setup again replaces the key, and the old one stops working.

inferam omp updates the provider for you before it starts. If you only run omp, it keeps working: Inferam answers a model it served before with the current one. Its context window stays the old model’s until you run inferam setup all (or inferam setup omp), so long sessions may not compact in time.

Terminal window
inferam setup omp --undo

Removes the inferam provider, with its key, from models.yml, and revokes the key. A file without it is left untouched.

Add a provider to ~/.omp/agent/models.yml:

providers:
inferam:
baseUrl: https://api.inferam.ai/v1
api: openai-completions
apiKey: INFERAM_API_KEY
models:
- id: DeepSeek-V4.1-Flash-1017
name: DeepSeek-V4.1-Flash-1017
reasoning: true
thinking: { mode: effort, efforts: [minimal, low, medium, high, xhigh, max], defaultLevel: high }
contextWindow: 262144
maxTokens: 32768

Then:

Terminal window
export INFERAM_API_KEY=ifr_...
omp --model inferam/DeepSeek-V4.1-Flash-1017

apiKey: INFERAM_API_KEY tells omp to read the key from that environment variable, so the key itself never goes in the file. (On Windows, set it with $env:INFERAM_API_KEY = "ifr_..." in PowerShell.) Without the thinking line, omp asks for its own default effort for this model, max.