🧭 feat: Add OpenRouter Prompt Cache Setting by danny-avila · Pull Request #13029 · danny-avila/LibreChat

danny-avila · 2026-05-08T19:53:39Z

Summary

I added OpenRouter prompt caching as a custom endpoint parameter and wired it through the OpenAI endpoint config path so LibreChat can pass promptCache to the agents OpenRouter client without leaking it to standard OpenAI requests.

Added OpenRouter parameter settings based on OpenAI plus the promptCache toggle.
Added OpenRouter schema parsing for defaultParamsEndpoint: openrouter, preserving promptCache for custom endpoint presets and compact conversations.
Passed promptCache into OpenRouter LLM config while respecting defaultParams, addParams, and dropParams.
Documented the OpenRouter custom endpoint example in librechat.example.yaml.
Added API and data-provider coverage for config, model option handling, and parser behavior.

Change Type

New feature (non-breaking change which adds functionality)
Documentation update

Testing

Ran npx jest src/endpoints/openai/llm.spec.ts src/endpoints/openai/config.spec.ts --runInBand --coverage=false from packages/api.
Ran npx jest specs/parsers.spec.ts --runInBand --coverage=false from packages/data-provider.
Ran npx tsc --noEmit --project packages/data-provider/tsconfig.json.
Ran focused ESLint on the touched API and data-provider files.
Tried npx tsc --noEmit --project packages/api/tsconfig.json; it fails on the existing Redis client type mismatch in packages/api/src/cache/cacheFactory.ts:47, outside this change.

Test Configuration:

Local worktree based on latest origin/dev at implementation time.

Checklist

My code adheres to this project's style guidelines
I have performed a self-review of my own code
I have made pertinent documentation changes
My changes do not introduce new warnings
I have written tests demonstrating that my changes are effective or that my feature works
Local unit tests pass with my changes

Copilot

Pull request overview

Adds an OpenRouter-specific promptCache toggle for custom endpoints, ensuring the setting is preserved through conversation parsing/compaction and only forwarded to the OpenRouter client configuration (not standard OpenAI requests).

Changes:

Added openRouterSchema and wired it into parseConvo / parseCompactConvo via defaultParamsEndpoint: openrouter support.
Exposed promptCache in OpenRouter parameter UI/config settings and validated it through custom endpoint config loading.
Passed promptCache through the OpenAI endpoint LLM config path only when useOpenRouter is true, with Jest coverage and updated librechat.example.yaml.

Reviewed changes

Copilot reviewed 9 out of 9 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
packages/data-provider/src/schemas.ts	Introduces `openRouterSchema` to preserve `promptCache` alongside OpenAI-like fields.
packages/data-provider/src/parsers.ts	Adds OpenRouter schema selection for `defaultParamsEndpoint`, including compact parsing support.
packages/data-provider/src/parameterSettings.ts	Adds OpenRouter parameter configuration including `promptCache` toggle.
packages/data-provider/specs/parsers.spec.ts	Tests that `promptCache` is preserved for OpenRouter custom endpoints in normal/compact parsing.
packages/api/src/types/openai.ts	Extends client options type to allow `promptCache` in LLM config.
packages/api/src/endpoints/openai/llm.ts	Applies `promptCache` via default/add/drop params and forwards it only for OpenRouter.
packages/api/src/endpoints/openai/llm.spec.ts	Adds tests ensuring `promptCache` is only emitted for OpenRouter and respects param precedence.
packages/api/src/endpoints/openai/config.spec.ts	Adds config-path coverage asserting `promptCache` flows from custom param definitions.
librechat.example.yaml	Documents `defaultParamsEndpoint: openrouter` and `promptCache` param definition example.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

@@ -28,10 +31,11 @@

 export type EndpointSchemaKey = EModelEndpoint;

-const endpointSchemas: Record<EndpointSchemaKey, EndpointSchema> = {
+const endpointSchemas: Record<string, EndpointSchema> = {
  [EModelEndpoint.openAI]: openAISchema,
  [EModelEndpoint.azureOpenAI]: openAISchema,
  [EModelEndpoint.custom]: openAISchema,
+  [Providers.OPENROUTER]: openRouterSchema,


 type CompactEndpointSchema =
  | typeof openAISchema
  | typeof compactAssistantSchema
  | typeof compactAgentsSchema
  | typeof compactGoogleSchema
+  | typeof openRouterSchema
  | typeof anthropicSchema
  | typeof bedrockInputSchema;

-const compactEndpointSchemas: Record<EndpointSchemaKey, CompactEndpointSchema> = {
+const compactEndpointSchemas: Record<string, CompactEndpointSchema> = {
  [EModelEndpoint.openAI]: openAISchema,
  [EModelEndpoint.azureOpenAI]: openAISchema,
  [EModelEndpoint.custom]: openAISchema,
+  [Providers.OPENROUTER]: openRouterSchema,
  [EModelEndpoint.assistants]: compactAssistantSchema,
  [EModelEndpoint.azureAssistants]: compactAssistantSchema,
  [EModelEndpoint.agents]: compactAgentsSchema,


github-actions · 2026-05-08T20:03:54Z

GitNexus: 🚀 deployed

The LibreChat-pr-13029 index is now live on the MCP server.
Deploy run

github-actions · 2026-05-08T21:12:14Z

GitNexus: 🚀 deployed

The LibreChat-pr-13029 index is now live on the MCP server.
Deploy run

danny-avila · 2026-05-08T21:53:18Z

@codex review

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 16335d6129

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

chatgpt-codex-connector · 2026-05-08T21:57:53Z

+  if (useOpenRouter && enablePromptCache === true) {
+    llmConfig.promptCache = true;


Honor OpenRouter promptCache for proxied endpoint names

promptCache is only forwarded when useOpenRouter is true, but useOpenRouter is derived in getOpenAIConfig from endpoint/baseURL string matching ("openrouter"), not from customParams.defaultParamsEndpoint. In setups that route OpenRouter through a custom domain or non-OpenRouter endpoint name, defaultParamsEndpoint: openrouter will preserve promptCache in parsing/UI yet this branch drops it before request construction, so the new prompt-cache toggle silently has no effect for those valid custom-endpoint configurations.

Useful? React with 👍 / 👎.

github-actions · 2026-05-08T22:36:37Z

GitNexus: 🚀 deployed

The LibreChat-pr-13029 index is now live on the MCP server.
Deploy run

github-actions · 2026-05-08T23:00:17Z

GitNexus: 🚀 deployed

The LibreChat-pr-13029 index is now live on the MCP server.
Deploy run

danny-avila · 2026-05-09T00:12:10Z

@codex review

chatgpt-codex-connector · 2026-05-09T00:17:15Z

Codex Review: Didn't find any major issues. More of your lovely PRs please.

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

github-actions · 2026-05-09T00:17:24Z

GitNexus: 🚀 deployed

The LibreChat-pr-13029 index is now live on the MCP server.
Deploy run

github-actions · 2026-05-09T00:32:03Z

GitNexus: 🚀 deployed

The LibreChat-pr-13029 index is now live on the MCP server.
Deploy run

github-actions · 2026-05-09T00:59:38Z

GitNexus: 🚀 deployed

The LibreChat-pr-13029 index is now live on the MCP server.
Deploy run

github-actions · 2026-05-09T01:14:47Z

GitNexus: 🚀 deployed

The LibreChat-pr-13029 index is now live on the MCP server.
Deploy run

danny-avila · 2026-05-09T02:40:38Z

@codex review

chatgpt-codex-connector · 2026-05-09T02:47:12Z

Codex Review: Didn't find any major issues. Can't wait for the next one!

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

github-actions · 2026-05-09T09:04:16Z

GitNexus: 🚀 deployed

The LibreChat-pr-13029 index is now live on the MCP server.
Deploy run

danny-avila · 2026-05-09T15:39:41Z

@codex review

chatgpt-codex-connector · 2026-05-09T15:44:01Z

Codex Review: Didn't find any major issues. Can't wait for the next one!

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

* feat: add OpenRouter prompt cache setting * fix: type OpenRouter schema lookup * fix: honor proxied OpenRouter prompt cache * refactor: flatten endpoint schema fallback * chore: Bump `@librechat/agents` to version 3.1.82 * fix: Default OpenRouter prompt cache params * test: Align OpenRouter config expectations * test: Update OpenRouter default cache expectation * fix: Align OpenRouter Detection * chore: Bump `@librechat/agents` to version 3.1.83 * docs: Remove OpenRouter prompt cache setup note * refactor: Use provider enum for OpenRouter defaults * style: Format OpenRouter defaults guard

Copilot AI review requested due to automatic review settings May 8, 2026 19:53

Copilot started reviewing on behalf of danny-avila May 8, 2026 19:54 View session

Copilot AI reviewed May 8, 2026

View reviewed changes

danny-avila marked this pull request as ready for review May 8, 2026 20:40

chatgpt-codex-connector Bot reviewed May 8, 2026

View reviewed changes

danny-avila added 4 commits May 8, 2026 20:06

feat: add OpenRouter prompt cache setting

1bd69fc

fix: type OpenRouter schema lookup

c7b7568

fix: honor proxied OpenRouter prompt cache

72a0341

refactor: flatten endpoint schema fallback

a905b4a

danny-avila force-pushed the danny-avila/openrouter-prompt-cache branch from 10867c6 to a905b4a Compare May 9, 2026 00:07

chore: Bump @librechat/agents to version 3.1.82

956efb8

danny-avila added 3 commits May 8, 2026 20:37

fix: Default OpenRouter prompt cache params

897b36a

test: Align OpenRouter config expectations

19b9f6c

test: Update OpenRouter default cache expectation

a80aafd

fix: Align OpenRouter Detection

eb2b645

chore: Bump @librechat/agents to version 3.1.83

1ea93d8

danny-avila added 2 commits May 9, 2026 11:36

docs: Remove OpenRouter prompt cache setup note

6df76f2

refactor: Use provider enum for OpenRouter defaults

e2ee400

style: Format OpenRouter defaults guard

e9d710d

danny-avila merged commit 8a654dc into dev May 9, 2026
15 checks passed

danny-avila deleted the danny-avila/openrouter-prompt-cache branch May 9, 2026 15:46

		if (useOpenRouter && enablePromptCache === true) {
		llmConfig.promptCache = true;

Uh oh!

Conversation

danny-avila commented May 8, 2026

Summary

Change Type

Testing

Test Configuration:

Checklist

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

github-actions Bot commented May 8, 2026

GitNexus: 🚀 deployed

Uh oh!

github-actions Bot commented May 8, 2026

GitNexus: 🚀 deployed

Uh oh!

danny-avila commented May 8, 2026

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

chatgpt-codex-connector Bot May 8, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented May 8, 2026

GitNexus: 🚀 deployed

Uh oh!

github-actions Bot commented May 8, 2026

GitNexus: 🚀 deployed

Uh oh!

danny-avila commented May 9, 2026

Uh oh!

chatgpt-codex-connector Bot commented May 9, 2026

Uh oh!

github-actions Bot commented May 9, 2026

GitNexus: 🚀 deployed

Uh oh!

github-actions Bot commented May 9, 2026

GitNexus: 🚀 deployed

Uh oh!

github-actions Bot commented May 9, 2026

GitNexus: 🚀 deployed

Uh oh!

github-actions Bot commented May 9, 2026

GitNexus: 🚀 deployed

Uh oh!

danny-avila commented May 9, 2026

Uh oh!

chatgpt-codex-connector Bot commented May 9, 2026

Uh oh!

github-actions Bot commented May 9, 2026

GitNexus: 🚀 deployed

Uh oh!

danny-avila commented May 9, 2026

Uh oh!

chatgpt-codex-connector Bot commented May 9, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants