PackagePackagetypeA library you install and call yourself.See every one →What this type is →by SAPby SAPoriginPublished by SAP itself. A visible facet, never an input to a reading — a dormant official plugin reads exactly like a dormant independent one.See every one →
@sap-ai-sdk/core
The shared internals that the packages of SAP's AI SDK build on, which its own readme says are not meant to be depended on directly.
LLM integrationCAPLLM integrationcapabilityCalling a language model from application code with the tenant, destination and credential handling already solved instead using an SDK call with a hand-rolled key.
Who is behind it
SAPSAPOrganisationMaintains this In the atlas: maintains 137, wrote 14@SAPmaintainerAnjanPersonContributes to this In the atlas: contributes to 7@anjan-poonachaSAP Cloud SDK BotPersonContributes to this In the atlas: contributes to 11@bot-sdk-jsBrigittaK307PersonContributes to this In the atlas: contributes to 12@BrigittaK307Charles DuboisPersonContributes to this In the atlas: contributes to 5@CharlesDuboisSAPChristoph MatthiesPersonContributes to this In the atlas: contributes to 5@ChristophMatthiesDavid KnaackPersonContributes to this In the atlas: contributes to 13, wrote 1@davidkna-sapDeeksha SinhaPersonContributes to this In the atlas: contributes to 12@deekshas8Guilherme DellagustinPersonContributes to this In the atlas: contributes to 27@dellagustin-sapemincihangeriPersonContributes to this In the atlas: contributes to 12@emincihangeriTian TengPersonContributes to this In the atlas: contributes to 5@Herrtian21 contributors
Before choosing this
crawled
Conditions
LicenceSAP Developer License Agreement 3.2 CAP§ 3from LLM integrationCovered SAP materials or a covered CAP customer application are used as training data or to develop or refine an AI model or system.
§ 3Artificial Intelligence Training. The agreement prohibits use of the covered materials and customer applications for training artificial-intelligence models or systems.
Calling a model for inference is not automatically AI training. Section 3 matters when the covered material or application participates in training, model development, refinement or the text-and-data-mining uses it names.
declared, Where the library runs once installed: inside its host (loaded into the CAP application, installed into the ABAP system, loaded by the UI5 application), in a program of your own, or only at build time.
Distributed through
npm
declared, How it is installed, as its documentation tells a user to get it., read npm install @sap-ai-sdk/core — @sap-ai-sdk/core — README, 2026-10-18
Connections1
Group by
NodeRelationConditionBasis
Functional1
LLM integrationLLM integrationcapabilityCalling a language model with tenant and credential handling solvedSAP and community, narrow by design
→provides
The shared internals of SAP's AI SDK, not meant to be depended on directly.
provides LLM integration
Functional, stated
For
The shared internals of SAP's AI SDK, not meant to be depended on directly.
Basis
“SAP Cloud SDK for AI is the official Software Development Kit (SDK) for SAP AI Core, SAP Generative AI Hub, and Orchestration Service.”SAP/ai-sdk-js — README, checked
stated
provides LLM integration
Functional, stated
For
The shared internals of SAP's AI SDK, not meant to be depended on directly.
Basis
“SAP Cloud SDK for AI is the official Software Development Kit (SDK) for SAP AI Core, SAP Generative AI Hub, and Orchestration Service.”SAP/ai-sdk-js — README, checked
Calling a language model with tenant and credential handling solved
This project is open to feature requests, bug reports and questions via GitHub issues.
Contribution and feedback are encouraged and always welcome.
For more information about how to contribute, the project structure, as well as additional contribution information, see our Contribution Guidelines.
0da56e2: [feat] Add LlmBatchModel type for models supported in LLM batch processing.
BatchCreateRequest.spec.model now uses this type for IDE autocomplete.
94b36e5: [Improvement] Added mistralai--mistral-medium, sap-rpt-1.5, sap-rpt-1.5-large, sap-rpt-1.6, sap-rpt-1.6-large, gpt-5.6-sol, gpt-5.6-terra, gpt-5.6-luna to the available model list.
Remove deprecated models gpt-4.1-nano (retirement date: 2026-10-14), gemini-2.5-flash (retirement date: 2026-10-16) and gemini-2.5-pro (retirement date: 2026-10-16).
4b2c014: [feat] Add gemini-3.5-flash-lite to the available model list.
Remove deprecated model mistralai--mistral-large-instruct (retirement date: 2026-09-30).
Patch Changes
3483acb: [fix] Set the default http agent socket timeout for AI Core requests to 1201 seconds (1201000 ms).
This overrides the Cloud SDK default of 5 seconds, which is too short for chat and streaming completions.
The value covers the orchestration server-side maximum of 1200 seconds plus 1 second of leeway.
Keep-alive is now disabled on these requests.
This avoids reusing stale sockets that a load balancer may have closed during the longer timeout.
Both settings apply only to the service-binding destination.
You can still override them with a custom destination's agentOptions or CustomRequestConfig.httpsAgent.
cca7edd: [Improvement] Add anthropic--claude-4.5-opus, anthropic--claude-4.8-opus, gemini-3.5-flash, gemini-embedding to the available model list.
cca7edd: [feat] Add GcpVertexAiEmbeddingModel type with gemini-embedding to @sap-ai-sdk/core.
Add GcpVertexAiEmbeddingModel to the EmbeddingModel union in @sap-ai-sdk/orchestration.
6542a2a: [feat] Add toReadableStream() method to SseStream.
The new method converts the asynchronous iterable stream to a pull-based ReadableStream of newline-delimited JSON, enabling composable transformations via the Web Streams API pipeThrough() pattern with native flow control.
0f10482: [feat] Add AzureOpenAiResponsesModel type for Azure OpenAI models that exclusively support the Responses API.
e944aa1: [Improvement] Added gpt-5.5 to the available model list.
Remove retired model anthropic--claude-4-opus — use anthropic--claude-4.7-opus instead.
Remove deprecated model o3-mini (retirement date: 2026-08-02).
Patch Changes
2faf4a8: [fix] Improve error message when server sends a non-JSON response during streaming.
cd3d8ed: [Improvement] Add gpt-5.2, anthropic--claude-4.5-opus, anthropic--claude-4.6-opus, anthropic--claude-4.6-sonnet and amazon-titan-embed-image to the available model list.
Removed deprecated gpt-4o model.
b12626b: [compat] Remove deprecated and retired models from model list.
Remove retired model anthropic--claude-3-sonnet from model list, use suggested replacement anthropic--claude-4.5-sonnet instead.
Remove retired model anthropic--claude-3-opus from model list.
Remove deprecated model gpt-4o-mini, use suggested replacement gpt-5-mini instead.
Remove deprecated models anthropic--claude-3.5-sonnet and anthropic--claude-3.7-sonnet.
Patch Changes
b11b00c: [feat] Advertise AbortSignal support for HTTP request cancellation.
This change adds typings, documentation and examples for using AbortSignal with the HTTP client to enable request cancellation.
56e9c3f: [Improvement] Combine 'ai-client-type' headers if a custom 'ai-client-type' header is set.
2e1d2c2: [Improvement] Added anthropic--claude-4.5-sonnet , anthropic--claude-4.5-haiku , gemini-2.5-flash-lite and sap-abap-1 to the available model list.
Removed deprecated models gemini-2.0-flash and gemini-2.0-flash-lite scheduled for retirement.
500c0dd: [Improvement] Add gpt-5,gpt-5-mini and gpt-5-nano to and remove gemini-1.5-flash, gemini-1.5-pro and ibm--granite-13b-chat from the available model list.
Patch Changes
9e1c43a: [Improvement] Add anthropic--claude-4-opus, anthropic--claude-4-sonnet, amazon--nova-premier, gemini-2.5-flash and gemini-2.5-pro to and remove mistralai--mixtral-8x7b-instruct-v01, meta--llama3.1-70b-instruct, nvidia--llama-3.2-nv-embedqa-1b, amazon--titan-embed-text, gpt-4, amazon--titan-text-express and amazon--titan-text-lite from the available model list.
a6ba3af: [Improvement] Add gemini-2.0-flash, gemini-2.0-flash-lite, anthropic--claude-3.7-sonnet and deepseek-ai--deepseek-r1 to and remove gemini-1.0-pro from the available model list.
bfed500: [Fixed Issue] Add missing cause in the error object for failing HTTP requests by using ErrorWithCause, providing more context for debugging.
[Compatibility Note] Due to the introduction of ErrorWithCause, AxiosError is now wrapped inside the cause property.
For example, use error.cause.response.data instead of error.response.data to access the error response from the server.
LLM integrationLLM integrationcapabilityCalling a language model from application code with the tenant, destination and credential handling already solved instead using an SDK call with a hand-rolled key.SAP and community, narrow by design
SAP AI CoreSAP AI CoreserviceSAP's AI service on BTP: the runtime for AI assets, and through the Generative AI Hub inside it the access to foundation models, the orchestration service with its templating, masking, filtering and grounding modules, the Prompt Registry and Document Grounding.sap official, btpAI runtime on BTP with the Generative AI Hub
cap-java/cds-aicap-java/cds-aipackageBridges CAP Java to SAP AI Core and adds AI-powered field recommendations to Fiori draft UIs, as two alpha Maven modules with mock fallbacks for local development.sap official, capBridge from CAP Java to SAP AI Core, with Fiori field recommendations
@cap-js/ai@cap-js/aicds-pluginOut-of-the-box value-help recommendations powered by SAP RPT-1, SAP AI Core as a standard CAP service, and an experimental local embedding model behind SQLite's vector functions.sap official, capValue-help recommendations and SAP AI Core as a CAP service
cap-llm-plugincap-llm-plugincds-pluginGenerative AI in a CAP application with the sensitive fields anonymised through SAP HANA before the prompt leaves, so the model never sees the confidential half.independent, capGenerative AI in CAP with fields anonymised through SAP HANA
@geertjanklaps/cds-toon-ai-protocol-adapter@geertjanklaps/cds-toon-ai-protocol-adaptercds-pluginA protocol adapter that serves CAP data as TOON — Token-Oriented Object Notation — instead of JSON, so a payload bound for a language model costs fewer tokens to send.independent, capProtocol adapter serving CAP data as TOON instead of JSON
@jerome-benoit/sap-ai-provider@jerome-benoit/sap-ai-providerpackageA provider for the Vercel AI SDK that speaks to SAP AI Core through its orchestration or foundation models API, so an application written against that SDK runs on SAP's models without being rewritten.independent, capVercel AI SDK provider for SAP AI Core
ui5-cap-ai-featuresui5-cap-ai-featurescds-pluginBoth halves of an AI text feature in one package: a CDS plugin serving an AIService over SAP AI Core's Orchestration Service, and the be.wl.ai UI5 control library that calls exactly that service.independent, capCDS plugin and UI5 controls for AI text features via SAP AI Core
@sap-ai-sdk/ai-api@sap-ai-sdk/ai-apipackageThe AI Core API surface of SAP's AI SDK, covering deployments, configurations and the rest of the control plane, as opposed to calling a model.sap official, capAI Core control plane API surface of SAP's AI SDK
@sap-ai-sdk/foundation-models@sap-ai-sdk/foundation-modelspackageThe package of SAP's AI SDK that calls foundation models through SAP AI Core and the Generative AI Hub, and the one an application actually reaches for.sap official, capClient for foundation models via SAP AI Core and Generative AI Hub
@sap-ai-sdk/orchestration@sap-ai-sdk/orchestrationpackageThe Orchestration Service client of SAP's AI SDK, which applies templating, grounding, filtering and masking around a model call rather than in the application.sap official, capOrchestration Service client for templating, grounding and masking
@sap-ai-sdk/prompt-registry@sap-ai-sdk/prompt-registrypackageA package for keeping prompts in SAP's prompt registry rather than in the codebase, so changing a prompt is not a deployment.sap official, capClient for prompts kept in SAP's prompt registry
@saptarishi/cds-plugin-llm@saptarishi/cds-plugin-llmcds-pluginA CAP service kind that puts eleven model providers behind one chat, stream, embed and batch interface, so which backend a service calls is a profile in package.json rather than a change to the handler.independent, capCAP service kind putting eleven model providers behind one API
@shahid.la/cds-db-nlquery-mcp@shahid.la/cds-db-nlquery-mcpmcp-serverMCP server that answers plain-English questions against the entities in db/schema.cds — it discovers the schema, generates CQN-backed SQL with real joins, and returns rows.independent, capMCP server answering plain-English questions against a CDS schema
