Repository navigation
Add Requesty as an OpenAI-compatible AI provider - #1720
Open
Thibaultjaigu wants to merge 2 commits into
Open
Thibaultjaigu wants to merge 2 commits into
Thibaultjaigu wants to merge 2 commits into
Conversation
Add Requesty, an OpenAI-compatible LLM router that exposes upstream models via provider/model naming (the same convention as OpenRouter). Mirrors the existing dedicated OpenAI-compatible providers (e.g. Z.ai): a RequestyProvider extending OpenAICompatibleProvider with base URL https://router.requesty.ai/v1 and REQUESTY_API_KEY auth. Wires it through the provider index, the PROVIDERS registry in ai-services-unified.js, the api-key map in config-manager.js, the VALIDATED_PROVIDERS / CUSTOM_PROVIDERS constants in tm-core, and adds a requesty model block to supported-models.json. Documents it in the README, docs/models.md, docs/configuration.md, llms-install.md, and env examples. Signed-off-by: Thibault Jaigu <thibault.jaigu@gmail.com>
🦋 Changeset detectedLatest commit: ebaa1c4 The changes in this PR will be included in the next version bump. This PR includes changesets to release 1 package
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
Contributor
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Advanced Run ID: 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
Included review availability: Your plan provides up to 10 included reviews per hour; 9 remain after this review. 📝 WalkthroughWalkthroughRequesty is added as an OpenAI-compatible provider with API-key configuration, runtime registration, supported model metadata, documentation, and a Changesets release note. ChangesRequesty provider support
Priority: ⬇️ Low Estimated code review effort: 3 (Moderate) | ~20 minutes Change: Feature Sequence Diagram(s)sequenceDiagram
participant UnifiedService
participant RequestyProvider
participant RequestyAPI
UnifiedService->>RequestyProvider: Resolve provider key requesty
RequestyProvider->>RequestyAPI: Send OpenAI-compatible request
RequestyAPI-->>RequestyProvider: Return model response
Merge Risk: ⚪ Minimal · up to The reviewed change adds the Requesty API-key status branch without an established merge-blocking risk. 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@scripts/modules/config-manager.js`:
- Line 903: Update the MCP provider switch in getMcpApiKeyStatus to add a
requesty case that uses the existing REQUESTY_API_KEY mapping and isApiKeySet
handling, preventing it from falling through to unknown-provider handling.
Preserve the behavior of all other providers.
In `@src/ai-providers/requesty.js`:
- Around line 15-24: The Requesty provider currently uses the class-based
architecture while the coding guidelines require legacy function-based
providers. In src/ai-providers/requesty.js lines 15-24, replace RequestyProvider
with standalone generateRequestyText, streamRequestyText, and
generateRequestyObject wrappers using Vercel AI SDK adapters and imports from ai
and the log utility; in scripts/modules/ai-services-unified.js line 75, register
those functions in PROVIDER_FUNCTIONS instead of PROVIDERS. If class-based
providers are now intentional, update the coding guidelines instead and preserve
the corresponding registry architecture.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: 9a23afc0-0888-4e51-b26c-05a64f3a2bab
📒 Files selected for processing (13)
.changeset/requesty-provider.md.env.exampleREADME.mdassets/env.exampledocs/configuration.mddocs/models.mdllms-install.mdpackages/tm-core/src/common/constants/providers.tsscripts/modules/ai-services-unified.jsscripts/modules/config-manager.jsscripts/modules/supported-models.jsonsrc/ai-providers/index.jssrc/ai-providers/requesty.js
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add Requesty as an OpenAI-compatible AI provider
This adds Requesty as a supported AI provider by mirroring the existing dedicated OpenAI-compatible providers.
Requesty is an OpenAI-compatible LLM router that exposes 400+ upstream models behind a single endpoint using the
provider/modelnaming convention (the same convention as OpenRouter), so it fits the existing provider architecture directly.Changes
src/ai-providers/requesty.js,RequestyProviderextendingOpenAICompatibleProvider(same pattern as the existingZ.aiprovider), base URLhttps://router.requesty.ai/v1,REQUESTY_API_KEYauth.src/ai-providers/index.js: exportRequestyProvider.scripts/modules/ai-services-unified.js: registerrequestyin thePROVIDERSmap.scripts/modules/config-manager.js: addrequesty: 'REQUESTY_API_KEY'to the API-key map.packages/tm-core/src/common/constants/providers.ts: addrequestytoVALIDATED_PROVIDERSandCUSTOM_PROVIDERS.scripts/modules/supported-models.json: add arequestymodel block (mirrors theopenrouterschema exactly) withopenai/gpt-4o-mini,openai/gpt-4o,openai/gpt-4.1-mini,anthropic/claude-sonnet-4-5,google/gemini-2.5-flash,deepseek/deepseek-chat.README.md,docs/models.md,docs/configuration.md,llms-install.md,assets/env.example,.env.example.Testing
npm run turbo:typecheck: passes (9/9 tasks, includes the full tm-core + extension compile).supported-models.jsonvalidated as JSON; therequestyblock has the same key shape as theopenrouterblock.openai/gpt-4o-minireturned the expected completion with HTTP 200.Note on the OpenRouter provider: it uses the OpenRouter-specific
@openrouter/ai-sdk-providerSDK, which isn't appropriate for a different endpoint, so Requesty is implemented via the repo's ownOpenAICompatibleProviderbase, the established form here for dedicated OpenAI-compatible brands. The OpenRouter-specific live model-fetching UI inapps/cli/src/commands/models/*(which hitsopenrouter.ai/api) was intentionally left untouched; Requesty models are served through the committedsupported-models.jsonlist.Config: get a key at https://app.requesty.ai/api-keys, model list at https://app.requesty.ai/router/list, docs at https://docs.requesty.ai.
I work at Requesty. This mirrors the existing OpenRouter provider as closely as possible. Happy to adjust or close it if it's not a fit.
Summary by CodeRabbit
New Features
REQUESTY_API_KEY.provider/modelnames.Documentation