feat: add endpoint for fetching unpriced models

docs/api/openapi.yml

@@ -4889,6 +4889,19 @@ components:
       - offset
       - limit
       type: object
+    LlmpricingruletypesGettableUnmappedModels:
+      properties:
+        items:
+          items:
+            $ref: '#/components/schemas/LlmpricingruletypesUnmappedModel'
+          nullable: true
+          type: array
+        total:
+          type: integer
+      required:
+      - items
+      - total
+      type: object
     LlmpricingruletypesLLMPricingCacheCosts:
       properties:
         mode:
@@ -4978,6 +4991,19 @@ components:
         type: string
       nullable: true
       type: array
+    LlmpricingruletypesUnmappedModel:
+      properties:
+        modelName:
+          type: string
+        provider:
+          type: string
+        spanCount:
+          minimum: 0
+          type: integer
+      required:
+      - modelName
+      - spanCount
+      type: object
     LlmpricingruletypesUpdatableLLMPricingRule:
       properties:
         enabled:
@@ -10451,6 +10477,60 @@ paths:
       summary: Get a pricing rule
       tags:
       - llmpricingrules
+  /api/v1/llm_pricing_rules/unmapped_models:
+    get:
+      deprecated: false
+      description: Returns models seen in the last hour of trace data (gen_ai.request.model)
+        that no pricing rule pattern matches, so the user can add them to an existing
+        rule or create a new one.
+      operationId: ListUnmappedLLMModels
+      responses:
+        "200":
+          content:
+            application/json:
+              schema:
+                properties:
+                  data:
+                    $ref: '#/components/schemas/LlmpricingruletypesGettableUnmappedModels'
+                  status:
+                    type: string
+                required:
+                - status
+                - data
+                type: object
+          description: OK
+        "400":
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/RenderErrorResponse'
+          description: Bad Request
+        "401":
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/RenderErrorResponse'
+          description: Unauthorized
+        "403":
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/RenderErrorResponse'
+          description: Forbidden
+        "500":
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/RenderErrorResponse'
+          description: Internal Server Error
+      security:
+      - api_key:
+        - VIEWER
+      - tokenizer:
+        - VIEWER
+      summary: List unmapped models
+      tags:
+      - llmpricingrules
   /api/v1/logs/promote_paths:
     get:
       deprecated: false

frontend/src/api/generated/services/llmpricingrules/index.ts

@@ -23,6 +23,7 @@ import type {
 	GetLLMPricingRulePathParameters,
 	ListLLMPricingRules200,
 	ListLLMPricingRulesParams,
+	ListUnmappedLLMModels200,
 	LlmpricingruletypesUpdatableLLMPricingRulesDTO,
 	RenderErrorResponseDTO,
 } from '../sigNoz.schemas';
@@ -393,3 +394,87 @@ export const invalidateGetLLMPricingRule = async (
 
 	return queryClient;
 };
+
+/**
+ * Returns models seen in the last hour of trace data (gen_ai.request.model) that no pricing rule pattern matches, so the user can add them to an existing rule or create a new one.
+ * @summary List unmapped models
+ */
+export const listUnmappedLLMModels = (signal?: AbortSignal) => {
+	return GeneratedAPIInstance<ListUnmappedLLMModels200>({
+		url: `/api/v1/llm_pricing_rules/unmapped_models`,
+		method: 'GET',
+		signal,
+	});
+};
+
+export const getListUnmappedLLMModelsQueryKey = () => {
+	return [`/api/v1/llm_pricing_rules/unmapped_models`] as const;
+};
+
+export const getListUnmappedLLMModelsQueryOptions = <
+	TData = Awaited<ReturnType<typeof listUnmappedLLMModels>>,
+	TError = ErrorType<RenderErrorResponseDTO>,
+>(options?: {
+	query?: UseQueryOptions<
+		Awaited<ReturnType<typeof listUnmappedLLMModels>>,
+		TError,
+		TData
+	>;
+}) => {
+	const { query: queryOptions } = options ?? {};
+
+	const queryKey = queryOptions?.queryKey ?? getListUnmappedLLMModelsQueryKey();
+
+	const queryFn: QueryFunction<
+		Awaited<ReturnType<typeof listUnmappedLLMModels>>
+	> = ({ signal }) => listUnmappedLLMModels(signal);
+
+	return { queryKey, queryFn, ...queryOptions } as UseQueryOptions<
+		Awaited<ReturnType<typeof listUnmappedLLMModels>>,
+		TError,
+		TData
+	> & { queryKey: QueryKey };
+};
+
+export type ListUnmappedLLMModelsQueryResult = NonNullable<
+	Awaited<ReturnType<typeof listUnmappedLLMModels>>
+>;
+export type ListUnmappedLLMModelsQueryError = ErrorType<RenderErrorResponseDTO>;
+
+/**
+ * @summary List unmapped models
+ */
+
+export function useListUnmappedLLMModels<
+	TData = Awaited<ReturnType<typeof listUnmappedLLMModels>>,
+	TError = ErrorType<RenderErrorResponseDTO>,
+>(options?: {
+	query?: UseQueryOptions<
+		Awaited<ReturnType<typeof listUnmappedLLMModels>>,
+		TError,
+		TData
+	>;
+}): UseQueryResult<TData, TError> & { queryKey: QueryKey } {
+	const queryOptions = getListUnmappedLLMModelsQueryOptions(options);
+
+	const query = useQuery(queryOptions) as UseQueryResult<TData, TError> & {
+		queryKey: QueryKey;
+	};
+
+	return { ...query, queryKey: queryOptions.queryKey };
+}
+
+/**
+ * @summary List unmapped models
+ */
+export const invalidateListUnmappedLLMModels = async (
+	queryClient: QueryClient,
+	options?: InvalidateOptions,
+): Promise<QueryClient> => {
+	await queryClient.invalidateQueries(
+		{ queryKey: getListUnmappedLLMModelsQueryKey() },
+		options,
+	);
+
+	return queryClient;
+};

frontend/src/api/generated/services/sigNoz.schemas.ts

@@ -6537,6 +6537,33 @@ export interface LlmpricingruletypesGettablePricingRulesDTO {
 	total: number;
 }
 
+export interface LlmpricingruletypesUnmappedModelDTO {
+	/**
+	 * @type string
+	 */
+	modelName: string;
+	/**
+	 * @type string
+	 */
+	provider?: string;
+	/**
+	 * @type integer
+	 * @minimum 0
+	 */
+	spanCount: number;
+}
+
+export interface LlmpricingruletypesGettableUnmappedModelsDTO {
+	/**
+	 * @type array,null
+	 */
+	items: LlmpricingruletypesUnmappedModelDTO[] | null;
+	/**
+	 * @type integer
+	 */
+	total: number;
+}
+
 export interface LlmpricingruletypesUpdatableLLMPricingRuleDTO {
 	/**
 	 * @type boolean
@@ -9474,6 +9501,14 @@ export type GetLLMPricingRule200 = {
 	status: string;
 };
 
+export type ListUnmappedLLMModels200 = {
+	data: LlmpricingruletypesGettableUnmappedModelsDTO;
+	/**
+	 * @type string
+	 */
+	status: string;
+};
+
 export type ListPromotedAndIndexedPaths200 = {
 	/**
 	 * @type array,null

frontend/src/container/MetricsApplication/Tabs/Overview/ServiceOverview.tsx

@@ -1,6 +1,5 @@
 import { useMemo } from 'react';
 import { useParams } from 'react-router-dom';
-import { Skeleton } from 'antd';
 import { ENTITY_VERSION_V4 } from 'constants/app';
 import { FeatureKeys } from 'constants/features';
 import { PANEL_TYPES } from 'constants/queryBuilder';
@@ -124,24 +123,14 @@ function ServiceOverview({
 			/>
 			<Card data-testid="service_latency">
 				<GraphContainer>
-					{topLevelOperationsIsLoading && (
-						<Skeleton
-							style={{
-								height: '100%',
-								padding: '16px',
-							}}
-						/>
-					)}
-					{!topLevelOperationsIsLoading && (
-						<Graph
-							onDragSelect={onDragSelect}
-							widget={latencyWidget}
-							onClickHandler={handleGraphClick('Service')}
-							isQueryEnabled={isQueryEnabled}
-							version={ENTITY_VERSION_V4}
-							enableDrillDown={SERVICE_DETAIL_DRILLDOWN_ENABLED}
-						/>
-					)}
+					<Graph
+						onDragSelect={onDragSelect}
+						widget={latencyWidget}
+						onClickHandler={handleGraphClick('Service')}
+						isQueryEnabled={isQueryEnabled}
+						version={ENTITY_VERSION_V4}
+						enableDrillDown={SERVICE_DETAIL_DRILLDOWN_ENABLED}
+					/>
 				</GraphContainer>
 			</Card>
 		</>

frontend/src/container/MetricsApplication/Tabs/Overview/TopLevelOperations.tsx

@@ -1,4 +1,3 @@
-import { Skeleton } from 'antd';
 import { Typography } from '@signozhq/ui/typography';
 import axios from 'axios';
 import { SOMETHING_WENT_WRONG } from 'constants/api';
@@ -29,24 +28,14 @@ function TopLevelOperation({
 				</Typography>
 			) : (
 				<GraphContainer>
-					{topLevelOperationsIsLoading && (
-						<Skeleton
-							style={{
-								height: '100%',
-								padding: '16px',
-							}}
-						/>
-					)}
-					{!topLevelOperationsIsLoading && (
-						<Graph
-							widget={widget}
-							onClickHandler={handleGraphClick(opName)}
-							onDragSelect={onDragSelect}
-							isQueryEnabled={!topLevelOperationsIsLoading}
-							version={ENTITY_VERSION_V4}
-							enableDrillDown={SERVICE_DETAIL_DRILLDOWN_ENABLED}
-						/>
-					)}
+					<Graph
+						widget={widget}
+						onClickHandler={handleGraphClick(opName)}
+						onDragSelect={onDragSelect}
+						isQueryEnabled={!topLevelOperationsIsLoading}
+						version={ENTITY_VERSION_V4}
+						enableDrillDown={SERVICE_DETAIL_DRILLDOWN_ENABLED}
+					/>
 				</GraphContainer>
 			)}
 		</Card>

pkg/apiserver/signozapiserver/llmpricingrule.go

@@ -49,6 +49,26 @@ func (provider *provider) addLLMPricingRuleRoutes(router *mux.Router) error {
 		return err
 	}
 
+	if err := router.Handle("/api/v1/llm_pricing_rules/unmapped_models", handler.New(
+		provider.authzMiddleware.ViewAccess(provider.llmPricingRuleHandler.ListUnmappedModels),
+		handler.OpenAPIDef{
+			ID:                  "ListUnmappedLLMModels",
+			Tags:                []string{"llmpricingrules"},
+			Summary:             "List unmapped models",
+			Description:         "Returns models seen in the last hour of trace data (gen_ai.request.model) that no pricing rule pattern matches, so the user can add them to an existing rule or create a new one.",
+			Request:             nil,
+			RequestContentType:  "",
+			Response:            new(llmpricingruletypes.GettableUnmappedModels),
+			ResponseContentType: "application/json",
+			SuccessStatusCode:   http.StatusOK,
+			ErrorStatusCodes:    []int{http.StatusBadRequest},
+			Deprecated:          false,
+			SecuritySchemes:     newSecuritySchemes(types.RoleViewer),
+		},
+	)).Methods(http.MethodGet).GetError(); err != nil {
+		return err
+	}
+
 	if err := router.Handle("/api/v1/llm_pricing_rules/{id}", handler.New(
 		provider.authzMiddleware.ViewAccess(provider.llmPricingRuleHandler.Get),
 		handler.OpenAPIDef{

pkg/modules/cloudintegration/config.go

@@ -22,7 +22,7 @@ func newConfig() factory.Config {
 		Agent: AgentConfig{
 			// we will maintain the latest version of cloud integration agent from here,
 			// till we automate it externally or figure out a way to validate it.
-			Version: "v0.0.12",
+			Version: "v0.0.13",
 		},
 	}
 }

pkg/modules/llmpricingrule/impllmpricingrule/handler.go

@@ -118,6 +118,28 @@ func (h *handler) CreateOrUpdate(rw http.ResponseWriter, r *http.Request) {
 	render.Success(rw, http.StatusNoContent, nil)
 }
 
+// ListUnmappedModels handles GET /api/v1/llm_pricing_rules/unmapped_models.
+func (h *handler) ListUnmappedModels(rw http.ResponseWriter, r *http.Request) {
+	ctx, cancel := context.WithTimeout(r.Context(), 30*time.Second)
+	defer cancel()
+
+	claims, err := authtypes.ClaimsFromContext(ctx)
+	if err != nil {
+		render.Error(rw, err)
+		return
+	}
+
+	orgID := valuer.MustNewUUID(claims.OrgID)
+
+	models, err := h.module.ListUnmappedModels(ctx, orgID)
+	if err != nil {
+		render.Error(rw, err)
+		return
+	}
+
+	render.Success(rw, http.StatusOK, llmpricingruletypes.NewGettableUnmappedModels(models))
+}
+
 // Delete handles DELETE /api/v1/llm_pricing_rules/{id}.
 func (h *handler) Delete(rw http.ResponseWriter, r *http.Request) {
 	ctx, cancel := context.WithTimeout(r.Context(), 10*time.Second)

pkg/modules/llmpricingrule/impllmpricingrule/module.go

@@ -3,22 +3,30 @@ package impllmpricingrule
 import (
 	"context"
 	"encoding/json"
+	"fmt"
 	"time"
 
 	"github.com/SigNoz/signoz/pkg/errors"
 	"github.com/SigNoz/signoz/pkg/modules/llmpricingrule"
+	"github.com/SigNoz/signoz/pkg/querier"
 	"github.com/SigNoz/signoz/pkg/query-service/agentConf"
 	"github.com/SigNoz/signoz/pkg/types/llmpricingruletypes"
 	"github.com/SigNoz/signoz/pkg/types/opamptypes"
+	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
+	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
 	"github.com/SigNoz/signoz/pkg/valuer"
 )
 
+// unmappedModelsLookback is the trace data window scanned to discover models in use.
+const unmappedModelsLookback = time.Hour
+
 type module struct {
-	store llmpricingruletypes.Store
+	store   llmpricingruletypes.Store
+	querier querier.Querier
 }
 
-func NewModule(store llmpricingruletypes.Store) llmpricingrule.Module {
-	return &module{store: store}
+func NewModule(store llmpricingruletypes.Store, querier querier.Querier) llmpricingrule.Module {
+	return &module{store: store, querier: querier}
 }
 
 func (module *module) List(ctx context.Context, orgID valuer.UUID, offset, limit int) ([]*llmpricingruletypes.LLMPricingRule, int, error) {
@@ -29,6 +37,28 @@ func (module *module) Get(ctx context.Context, orgID valuer.UUID, id valuer.UUID
 	return module.store.Get(ctx, orgID, id)
 }
 
+// ListUnmappedModels discovers the models present in the last hour of trace data
+// (gen_ai.request.model) and returns the ones that no pricing rule pattern matches.
+func (module *module) ListUnmappedModels(ctx context.Context, orgID valuer.UUID) ([]*llmpricingruletypes.UnmappedModel, error) {
+	models, err := module.discoverModels(ctx, orgID)
+	if err != nil {
+		return nil, err
+	}
+
+	rules, _, err := module.store.List(ctx, orgID, 0, 10000)
+	if err != nil {
+		return nil, err
+	}
+
+	unmapped := make([]*llmpricingruletypes.UnmappedModel, 0, len(models))
+	for _, m := range models {
+		if !llmpricingruletypes.ModelMatchesAnyRule(m.ModelName, rules) {
+			unmapped = append(unmapped, m)
+		}
+	}
+	return unmapped, nil
+}
+
 // CreateOrUpdate applies a batch of pricing rule changes:
 //   - ID set       → match by id, overwrite fields.
 //   - SourceID set → match by source_id; if found overwrite, else insert.
@@ -135,3 +165,108 @@ func (module *module) findExisting(ctx context.Context, orgID valuer.UUID, u *ll
 		return nil, errors.Newf(errors.TypeNotFound, llmpricingruletypes.ErrCodePricingRuleNotFound, "rule has neither id nor sourceId")
 	}
 }
+
+// discoverModels runs a QBv5 traces aggregation grouped by gen_ai.request.model
+// over the lookback window and returns each distinct model with its span count.
+func (module *module) discoverModels(ctx context.Context, orgID valuer.UUID) ([]*llmpricingruletypes.UnmappedModel, error) {
+	now := time.Now()
+	req := &qbtypes.QueryRangeRequest{
+		Start:       uint64(now.Add(-unmappedModelsLookback).UnixMilli()),
+		End:         uint64(now.UnixMilli()),
+		RequestType: qbtypes.RequestTypeScalar,
+		CompositeQuery: qbtypes.CompositeQuery{
+			Queries: []qbtypes.QueryEnvelope{
+				{
+					Type: qbtypes.QueryTypeBuilder,
+					Spec: qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+						Name:   "A",
+						Signal: telemetrytypes.SignalTraces,
+						Filter: &qbtypes.Filter{Expression: fmt.Sprintf("%s EXISTS", llmpricingruletypes.GenAIRequestModel)},
+						Aggregations: []qbtypes.TraceAggregation{
+							{Expression: "count()", Alias: "spanCount"},
+						},
+						GroupBy: []qbtypes.GroupByKey{
+							{TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{
+								Name:          llmpricingruletypes.GenAIRequestModel,
+								FieldContext:  telemetrytypes.FieldContextSpan,
+								FieldDataType: telemetrytypes.FieldDataTypeString,
+							}},
+							{TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{
+								Name:          llmpricingruletypes.GenAIProviderName,
+								FieldContext:  telemetrytypes.FieldContextSpan,
+								FieldDataType: telemetrytypes.FieldDataTypeString,
+							}},
+						},
+						Limit: 1000,
+					},
+				},
+			},
+		},
+	}
+
+	resp, err := module.querier.QueryRange(ctx, orgID, req)
+	if err != nil {
+		return nil, err
+	}
+
+	return parseModels(resp), nil
+}
+
+// parseModels extracts the grouped model names and their span counts from a scalar response.
+func parseModels(resp *qbtypes.QueryRangeResponse) []*llmpricingruletypes.UnmappedModel {
+	if resp == nil || len(resp.Data.Results) == 0 {
+		return nil
+	}
+	sd, ok := resp.Data.Results[0].(*qbtypes.ScalarData)
+	if !ok || sd == nil {
+		return nil
+	}
+
+	modelIdx, providerIdx, countIdx := -1, -1, -1
+	for i, c := range sd.Columns {
+		switch c.Type {
+		case qbtypes.ColumnTypeGroup:
+			switch c.Name {
+			case llmpricingruletypes.GenAIRequestModel:
+				modelIdx = i
+			case llmpricingruletypes.GenAIProviderName:
+				providerIdx = i
+			}
+		case qbtypes.ColumnTypeAggregation:
+			countIdx = i
+		}
+	}
+	if modelIdx == -1 {
+		return nil
+	}
+
+	models := make([]*llmpricingruletypes.UnmappedModel, 0, len(sd.Data))
+	for _, row := range sd.Data {
+		name, _ := row[modelIdx].(string)
+		if name == "" {
+			continue
+		}
+		provider := ""
+		if providerIdx != -1 {
+			provider, _ = row[providerIdx].(string)
+		}
+		models = append(models, &llmpricingruletypes.UnmappedModel{ModelName: name, Provider: provider, SpanCount: toUint64(row, countIdx)})
+	}
+	return models
+}
+
+func toUint64(row []any, idx int) uint64 {
+	if idx < 0 || idx >= len(row) {
+		return 0
+	}
+	switch v := row[idx].(type) {
+	case uint64:
+		return v
+	case int64:
+		return uint64(v)
+	case float64:
+		return uint64(v)
+	default:
+		return 0
+	}
+}

pkg/modules/llmpricingrule/llmpricingrule.go

@@ -17,6 +17,7 @@ type Module interface {
 	Get(ctx context.Context, orgID valuer.UUID, id valuer.UUID) (*llmpricingruletypes.LLMPricingRule, error)
 	CreateOrUpdate(ctx context.Context, orgID valuer.UUID, userEmail string, rules []*llmpricingruletypes.UpdatableLLMPricingRule) (err error)
 	Delete(ctx context.Context, orgID, id valuer.UUID) error
+	ListUnmappedModels(ctx context.Context, orgID valuer.UUID) ([]*llmpricingruletypes.UnmappedModel, error)
 }
 
 // Handler defines the HTTP handler interface for pricing rule endpoints.
@@ -25,4 +26,5 @@ type Handler interface {
 	Get(rw http.ResponseWriter, r *http.Request)
 	CreateOrUpdate(rw http.ResponseWriter, r *http.Request)
 	Delete(rw http.ResponseWriter, r *http.Request)
+	ListUnmappedModels(rw http.ResponseWriter, r *http.Request)
 }

pkg/signoz/module.go

@@ -155,7 +155,7 @@ func NewModules(
 		CloudIntegration: cloudIntegrationModule,
 		TraceDetail:      impltracedetail.NewModule(impltracedetail.NewTraceStore(telemetryStore), providerSettings, config.TraceDetail),
 		SpanMapper:       implspanmapper.NewModule(implspanmapper.NewStore(sqlstore)),
-		LLMPricingRule:   impllmpricingrule.NewModule(impllmpricingrule.NewStore(sqlstore)),
+		LLMPricingRule:   impllmpricingrule.NewModule(impllmpricingrule.NewStore(sqlstore), querier),
 		Tag:              tagModule,
 	}
 }

pkg/types/llmpricingruletypes/pricing.go

@@ -3,6 +3,7 @@ package llmpricingruletypes
 import (
 	"database/sql/driver"
 	"encoding/json"
+	"path"
 	"time"
 
 	"github.com/SigNoz/signoz/pkg/errors"
@@ -16,6 +17,7 @@ const (
 	LLMCostFeatureType agentConf.AgentFeatureType = "llm_pricing"
 
 	GenAIRequestModel                  = "gen_ai.request.model"
+	GenAIProviderName                  = "gen_ai.provider.name"
 	GenAIUsageInputTokens              = "gen_ai.usage.input_tokens"
 	GenAIUsageOutputTokens             = "gen_ai.usage.output_tokens"
 	GenAIUsageCacheReadInputTokens     = "gen_ai.usage.cache_read.input_tokens"
@@ -136,6 +138,32 @@ type GettablePricingRules struct {
 	Limit  int                       `json:"limit"  required:"true"`
 }
 
+// UnmappedModel is a model observed in trace data (gen_ai.request.model) that
+// no pricing rule pattern matches, so no cost is being computed for it.
+type UnmappedModel struct {
+	ModelName string `json:"modelName" required:"true"`
+	Provider  string `json:"provider"`
+	SpanCount uint64 `json:"spanCount" required:"true"`
+}
+
+type GettableUnmappedModels struct {
+	Items []*UnmappedModel `json:"items" required:"true"`
+	Total int              `json:"total" required:"true"`
+}
+
+// ModelMatchesAnyRule reports whether model matches any rule's glob pattern,
+// mirroring the path.Match semantics the signozllmpricing OTel processor uses.
+func ModelMatchesAnyRule(model string, rules []*LLMPricingRule) bool {
+	for _, r := range rules {
+		for _, pattern := range r.ModelPattern {
+			if ok, err := path.Match(pattern, model); err == nil && ok {
+				return true
+			}
+		}
+	}
+	return false
+}
+
 func (LLMPricingRuleUnit) Enum() []any {
 	return []any{UnitPerMillionTokens}
 }
@@ -204,6 +232,13 @@ func NewGettableLLMPricingRulesFromLLMPricingRules(items []*LLMPricingRule, tota
 	}
 }
 
+func NewGettableUnmappedModels(items []*UnmappedModel) *GettableUnmappedModels {
+	return &GettableUnmappedModels{
+		Items: items,
+		Total: len(items),
+	}
+}
+
 func NewLLMPricingRuleFromUpdatable(u *UpdatableLLMPricingRule, orgID valuer.UUID, userEmail string, now time.Time) *LLMPricingRule {
 	isOverride := true
 	if u.IsOverride != nil {