feat(inference): add Nemotron 3 Nano Omni to CLOUD_MODEL_OPTIONS

[PicoNVIDIA]

Summary

Adds nvidia/nemotron-3-nano-omni-30b-a3b-reasoning (label: Nemotron 3 Nano Omni 30B) to the curated cloud model picker. Super 120B remains the default.

Motivation

The multimodal hermes-omni-demo cookbook in brevdev/nemoclaw-demos currently has to do a post-onboard openshell inference set --model private/nvidia/nemotron-3-nano-omni-30b-a3b-reasoning to switch the gateway from Super to Omni, because the wizard only exposes Super. The current cookbook frames this as "You picked Super 120B during onboarding because that's what the menu offers, but this cookbook needs Omni..." — that workaround is awkward and reviewers in brevdev/nemoclaw-demos#23 have called it out as an abuse of the installer.

Adding Omni here lets users select it during nemoclaw onboard directly and lets multimodal cookbooks drop the manual swap step entirely.

Test plan

• Existing inference-config.test.ts updated to include the new model id in the expected list — runs as part of npm test
• nemoclaw onboard --agent hermes shows Omni as option in the model picker
• Selecting it produces a sandbox with Model: private/nvidia/nemotron-3-nano-omni-30b-a3b-reasoning in nemoclaw <name> status

Files changed

• src/lib/inference-config.ts — one new entry in CLOUD_MODEL_OPTIONS
• src/lib/inference-config.test.ts — matching expected-list update

🤖 Generated with Claude Code

Summary by CodeRabbit

• New Features

  • NVIDIA Nemotron 3 Nano Omni 30B added to cloud model selection for users to choose.

• Tests

  • Automated tests updated to include the new cloud model option and to reflect adjusted selection ordering used in onboarding and default-model scenarios.

Signed-off-by: Aaron Erickson aerickson@nvidia.com

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[coderabbitai]

Note

Reviews paused

It looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the reviews.auto_review.auto_pause_after_reviewed_commits setting.

Use the following commands to manage reviews:

• @coderabbitai resume to resume automatic reviews.
• @coderabbitai review to trigger a single review.

Use the checkboxes below for quick actions:

• ▶️ Resume reviews
• 🔍 Trigger review

📝 Walkthrough

Walkthrough

Added nvidia/nemotron-3-nano-omni-30b-a3b-reasoning to CLOUD_MODEL_OPTIONS and updated tests and mocked prompt selections to account for the new option, adjusting index-based choices in related onboarding and prompt tests.

Changes

Cloud model options and tests

Layer / File(s)	Summary
Data Shape src/lib/inference-config.ts	Appended a new entry to CLOUD_MODEL_OPTIONS with id nvidia/nemotron-3-nano-omni-30b-a3b-reasoning and label Nemotron 3 Nano Omni 30B.
Tests — curated picker expectation src/lib/inference-config.test.ts	Updated test expectation to include the new model ID in the curated cloud model picker options.
Tests — default cloud model prompt src/lib/model-prompts.test.ts	Adjusted mocked prompt input from "5" to "6" so the test still returns the same expected default model choice (deepseek-ai/deepseek-v4-pro).
Tests — onboarding selections test/onboard-selection.test.ts	Updated mocked onboarding answers: DeepSeek V4 Pro selection index changed from 5→6, and the manual NVIDIA model flow index changed from 6→7 before entering custom/provider-model.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Poem

🐰 I hopped the list, a tiny tweak,
A Nemotron added, nimble and sleek.
Tests nudged one step to the right,
Prompts and onboarding set aright.
A small change, and all feels bright.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the main change: adding a new NVIDIA Nemotron model to the cloud model options configuration.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[wscurran]

✨ Thanks for submitting this PR that adds Nemotron 3 Nano Omni to the curated cloud model picker, allowing users to select it during onboard directly. This change aims to improve the user experience by eliminating the need for a manual model swap step in multimodal cookbooks.

---

Related open issues:

• #23 Make nemoclaw onboard robust when rerun after a failed install instead of requiring uninstall.sh as a workaround

[ericksoa]

Approved after fixing the NVIDIA Build model ID to nvidia/nemotron-3-nano-omni-30b-a3b-reasoning, resolving the main-branch conflict, updating the shifted model-picker tests, and verifying the focused/local test coverage plus the completed PR checks.