#9109: Models: add Orq provider
docs
cli
commands
agents
size: L
Cluster:
Model Provider Integrations
## Summary
- add Orq provider support with OpenAI-compatible router base URL
- add curated Orq model catalog (Anthropic 4.5, Gemini 2.5/3, GPT‑5/5.2, Groq, Cerebras)
- wire Orq into onboarding, non‑interactive flags, auth choice, and model status probing
- document Orq provider and onboarding steps
## Details
- Models
- added an Orq catalog with ids, names, context windows, max tokens, costs, and input modalities
- included both google-ai (Gemini API) and google (Vertex) variants for Gemini
- set Orq default model to `orq/openai/gpt-5.2`
- Onboarding
- new `--orq-api-key` flag for non‑interactive onboarding
- Orq now appears in the auth choice menu and preferred‑provider mapping
- onboarding applies `models.providers.orq` with base URL + catalog and sets default model
- Auth
- supports `ORQ_API_KEY` env fallback for provider auth
- CLI
- `models status --probe --probe-provider orq` supported
## Testing
- pnpm test
## Notes
- Orq models are stored in `src/agents/orq-models.ts` to keep onboarding wiring lighter.
- Pricing is sourced from Orq’s model metadata; one model is currently `0` cost where pricing is unavailable.
<!-- greptile_comment -->
<h2>Greptile Overview</h2>
<h3>Greptile Summary</h3>
This PR adds a new `orq` model provider wired to the existing OpenAI-compatible router flow, including a curated Orq model catalog (`src/agents/orq-models.ts`), onboarding wiring for interactive and non-interactive flows (new `orq-api-key` auth choice + `--orq-api-key` flag), and docs under `docs/providers/orq.md` plus cross-links in provider indexes.
Codebase fit: the implementation follows the existing onboarding/provider pattern used by OpenRouter/Moonshot/etc by (1) teaching `resolveEnvApiKey` about `ORQ_API_KEY`, (2) adding an auth profile setter (`setOrqApiKey`) and auth-choice plumbing, and (3) adding `applyOrqConfig/applyOrqProviderConfig` to stamp provider defaults + model aliases into config.
<h3>Confidence Score: 3/5</h3>
- This PR is close to merge-ready but has test failures and one behavior issue to address.
- Confidence reduced because the added/modified tests contain deterministic assertion mismatches (env var name vs value, model id mismatch), and `applyOrqProviderConfig` currently overwrites existing user provider settings (baseUrl/api), which can silently discard valid configuration when onboarding is rerun. Aside from those, the changes follow existing provider/onboarding patterns.
- src/agents/models-config.fills-missing-provider-apikey-from-env-var.test.ts, src/commands/onboard-auth.config-core.ts
<!-- greptile_other_comments_section -->
<sub>(2/5) Greptile learns from your feedback when you react with thumbs up/down!</sub>
<!-- /greptile_comment -->
Most Similar PRs
#20965: feat: Add comprehensive model configuration and discovery for various…
by rodeok · 2026-02-19
79.0%
#8135: Adding Groq as a model provider
by FameDied · 2026-02-03
78.1%
#7570: fix: allow models from providers with auth profiles configured
by DonSqualo · 2026-02-03
77.1%
#7272: Models: add SiliconFlow provider
by qychen2001 · 2026-02-02
76.8%
#7418: feat (amazon-nova): add Amazon Nova 1P provider
by 5herlocked · 2026-02-02
76.1%
#15991: feat: add Novita AI provider support with dynamic model discovery
by Alex-wuhu · 2026-02-14
75.9%
#9822: fix: allow local/custom model providers for sub-agent inference
by stammtobias91 · 2026-02-05
75.6%
#13079: feat: Add OpenAI-compatible API option to CLI for self-hosted models
by MikeWang0316tw · 2026-02-10
75.4%
#13295: feat: add Eternal AI provider integration
by peterparkernho · 2026-02-10
75.3%
#16766: fix(model): apply provider baseUrl/headers override to registry-fou...
by dzianisv · 2026-02-15
75.0%