Same agent Smaller bill. The tools and infrastructure to get more work from your AI. Fewer tokens. Less compute. Lower cost.
TypeScript agent-harnessagent-harness-engineeringai-engineeringai-observabilityamd-gpuclaude-codeclickhousecodexcoding-agentcoding-agentscursorgenaigpu-monitoringllm-observabilityllmopsnvidia-smiobservabilityopen-sourceopenaiopentelemetry
feat(client,controller): add OrcaRouter as a first-class provider (#1493)
kuswardhanietidims-svgfc07f16341bb
diff --git a/docs/latest/openlit/costs/manage-models/overview.mdx b/docs/latest/openlit/costs/manage-models/overview.mdxindex 42110b8e..d86d08ed 100644--- a/docs/latest/openlit/costs/manage-models/overview.mdx+++ b/docs/latest/openlit/costs/manage-models/overview.mdx@@ -3,7 +3,7 @@ title: "Overview"description: "Manage models on the Costs page - edit LLM providers and per-model token prices used for AI cost tracking across traces, Otter, OpenGround, and the SDK"----**Manage models** is the **Manage models** tab on **Costs** (`/costs?tab=models`) under Monitor. Every connected ClickHouse database is seeded with **14 providers** and a large built-in model catalog (90+ models). You can edit any price, add custom providers and models, import/export pricing JSON, and point the OpenLIT SDK at a public pricing URL so ingest-time cost tracking stays in sync with the UI.+**Manage models** is the **Manage models** tab on **Costs** (`/costs?tab=models`) under Monitor. Every connected ClickHouse database is seeded with **16 providers** and a large built-in model catalog (90+ models). You can edit any price, add custom providers and models, import/export pricing JSON, and point the OpenLIT SDK at a public pricing URL so ingest-time cost tracking stays in sync with the UI.Those prices power [Configuration](/latest/openlit/costs/configuration) (Auto Pricing and manual recalculation), [Analytics](/latest/openlit/costs/analytics), cost columns in telemetry, [OpenGround](/latest/openlit/prompts-experiments/openground/overview) comparisons, and Otter's message cost estimates.@@ -26,7 +26,7 @@ Open **Monitor → Costs → Manage models**. (`/manage-models` redirects here.)## Built-in providers-Out of the box you get: OpenAI, Anthropic, Google AI, Mistral AI, Groq, Perplexity, Azure OpenAI, Cohere, Together AI, Fireworks AI, DeepSeek, xAI, Hugging Face, and Replicate. Model lists change as the catalog is updated - treat Manage models itself as the source of truth for what is installed in your database.+Out of the box you get: OpenAI, Anthropic, Google AI, Mistral AI, Groq, Perplexity, Azure OpenAI, Cohere, Together AI, Fireworks AI, DeepSeek, xAI, Hugging Face, Replicate, MiniMax, and OrcaRouter. Model lists change as the catalog is updated - treat Manage models itself as the source of truth for what is installed in your database.## API referencediff --git a/docs/latest/openlit/costs/manage-models/providers.mdx b/docs/latest/openlit/costs/manage-models/providers.mdxindex 1c5af372..82da4d89 100644--- a/docs/latest/openlit/costs/manage-models/providers.mdx+++ b/docs/latest/openlit/costs/manage-models/providers.mdx@@ -3,7 +3,7 @@ title: "Providers"description: "Add and edit LLM providers in Manage Models - Provider ID, display name, description, and Requires API Key (Vault)"----Providers are the top-level grouping for models on **Manage Models** (`/costs?tab=models`). OpenLIT ships 14 built-in providers; you can add your own for self-hosted or niche APIs.+Providers are the top-level grouping for models on **Manage Models** (`/costs?tab=models`). OpenLIT ships 16 built-in providers; you can add your own for self-hosted or niche APIs.## Add a providerdiff --git a/docs/latest/openlit/prompts-experiments/openground/overview.mdx b/docs/latest/openlit/prompts-experiments/openground/overview.mdxindex f256de00..6a0220b2 100644--- a/docs/latest/openlit/prompts-experiments/openground/overview.mdx+++ b/docs/latest/openlit/prompts-experiments/openground/overview.mdx@@ -41,7 +41,7 @@ Model and provider management itself - adding providers and editing pricing - liOpenGround runs the same prompt across multiple LLM providers and models at once, so you can compare cost, speed, and response quality side-by-side before choosing a model for production.</Accordion><Accordion title="How many providers are supported out of the box?">- 14: OpenAI, Anthropic, Google, Mistral, Groq, Perplexity, Azure, Cohere, Together AI, Fireworks, DeepSeek, xAI, Hugging Face, and Replicate. See [Run a Comparison](/latest/openlit/prompts-experiments/openground/run-a-comparison) for how to configure one.+ 16: OpenAI, Anthropic, Google, Mistral, Groq, Perplexity, Azure, Cohere, Together AI, Fireworks, DeepSeek, xAI, Hugging Face, Replicate, MiniMax, and OrcaRouter. See [Run a Comparison](/latest/openlit/prompts-experiments/openground/run-a-comparison) for how to configure one.</Accordion><Accordion title="Can I test a model that isn't built in?">Yes. When configuring a provider in OpenGround, switch to "Use Custom Model" and type any model ID directly, with its own input/output pricing. There's no field for a custom endpoint or base URL - only the model ID and its pricing are configurable.diff --git a/docs/latest/openlit/prompts-experiments/openground/run-a-comparison.mdx b/docs/latest/openlit/prompts-experiments/openground/run-a-comparison.mdxindex c08a9f00..52ee652a 100644--- a/docs/latest/openlit/prompts-experiments/openground/run-a-comparison.mdx+++ b/docs/latest/openlit/prompts-experiments/openground/run-a-comparison.mdx@@ -23,7 +23,7 @@ A grid of provider cards follows, each showing an avatar, name, model count, andAs soon as you've added at least one provider, a banner shows **"{n} Providers Selected"**.-OpenGround ships with **14 built-in providers**: OpenAI, Anthropic, Google, Mistral, Groq, Perplexity, Azure, Cohere, Together AI, Fireworks, DeepSeek, xAI, Hugging Face, and Replicate. These are the same providers you'd see on [Manage models](/latest/openlit/costs/manage-models/overview) under Costs - anything you add or configure there is immediately available here, and vice versa.+OpenGround ships with **16 built-in providers**: OpenAI, Anthropic, Google, Mistral, Groq, Perplexity, Azure, Cohere, Together AI, Fireworks, DeepSeek, xAI, Hugging Face, Replicate, MiniMax, and OrcaRouter. These are the same providers you'd see on [Manage models](/latest/openlit/costs/manage-models/overview) under Costs - anything you add or configure there is immediately available here, and vice versa.## Step 3: Provider settingsdiff --git a/openlit-controller/internal/engine/obi.go b/openlit-controller/internal/engine/obi.goindex 19cc0a05..0493add9 100644--- a/openlit-controller/internal/engine/obi.go+++ b/openlit-controller/internal/engine/obi.go@@ -392,11 +392,11 @@ func BuildInstrumentConfig(// OBI has native parsers only for OpenAI, Anthropic, Gemini, Qwen, and// Bedrock. Every other OpenAI-wire-compatible SaaS provider we discover- // (Cohere/Mistral/Groq/Deepseek/Together/Fireworks/Vercel/Vertex/Azure...)+ // (Cohere/Mistral/Groq/Deepseek/Together/Fireworks/OrcaRouter/Vercel/Vertex/Azure...)// is parsed by the OpenAI extractor, so enabling any of them enables OpenAI.openAICompatible := []string{"openai", "cohere", "mistral", "groq", "deepseek", "together",- "fireworks", "azure_inference", "azure_openai", "vercel_ai", "vertex_ai",+ "fireworks", "orcarouter", "azure_inference", "azure_openai", "vercel_ai", "vertex_ai",}for _, p := range openAICompatible {if enabledProviders[p] {diff --git a/openlit-controller/internal/scanner/bpf/llm_scanner.h b/openlit-controller/internal/scanner/bpf/llm_scanner.hindex b5cfd473..08ec7d75 100644--- a/openlit-controller/internal/scanner/bpf/llm_scanner.h+++ b/openlit-controller/internal/scanner/bpf/llm_scanner.h@@ -3,8 +3,8 @@/* Provider IDs are assigned and interpreted on the Go side (see* internal/scanner/types.go); these defines are informational. Built-in SaaS- * providers use 1-12 (+ Bedrock regions 20-26); all user-configured custom- * gateways collapse to a single PROVIDER_CUSTOM. */+ * providers use 1-12 and 14 (+ Bedrock regions 20-26); all user-configured+ * custom gateways collapse to a single PROVIDER_CUSTOM. */#define PROVIDER_OPENAI 1#define PROVIDER_ANTHROPIC 2#define PROVIDER_GEMINI 3@@ -15,6 +15,7 @@#define PROVIDER_TOGETHER 8#define PROVIDER_FIREWORKS 9#define PROVIDER_CUSTOM 13+#define PROVIDER_ORCAROUTER 14/* Key for the llm_endpoints map: destination IPv4 (network byte order) plus* destination port (host byte order). Keying on IP+port lets us matchdiff --git a/openlit-controller/internal/scanner/hosts.go b/openlit-controller/internal/scanner/hosts.goindex 8b59cf04..b77d17f8 100644--- a/openlit-controller/internal/scanner/hosts.go+++ b/openlit-controller/internal/scanner/hosts.go@@ -42,6 +42,7 @@ var builtinTargets = []llmTarget{{Host: "ai-gateway.vercel.sh", Port: 443, ProviderID: 10}, // PROVIDER_VERCEL_AI{Host: "aiplatform.googleapis.com", Port: 443, ProviderID: 11}, // PROVIDER_VERTEX_AI{Host: "models.inference.ai.azure.com", Port: 443, ProviderID: 12}, // PROVIDER_AZURE_INFERENCE (GitHub Models)+ {Host: "api.orcarouter.ai", Port: 443, ProviderID: 14}, // PROVIDER_ORCAROUTER}// Well-known Bedrock regions. New regions are detected dynamically ifdiff --git a/openlit-controller/internal/scanner/types.go b/openlit-controller/internal/scanner/types.goindex 48f3fae3..add30a82 100644--- a/openlit-controller/internal/scanner/types.go+++ b/openlit-controller/internal/scanner/types.go@@ -20,6 +20,7 @@ var providerNames = map[uint8]string{11: "vertex_ai",12: "azure_inference",13: "custom",+ 14: "orcarouter",// IDs 20-26 are Bedrock regions (assigned dynamically in hosts.go init)20: "bedrock",21: "bedrock",diff --git a/src/client/src/__tests__/lib/platform/chat/stream.test.ts b/src/client/src/__tests__/lib/platform/chat/stream.test.tsindex 7b49524b..790f21c8 100644--- a/src/client/src/__tests__/lib/platform/chat/stream.test.ts+++ b/src/client/src/__tests__/lib/platform/chat/stream.test.ts@@ -105,7 +105,7 @@ describe('getModelInstance', () => {const providers = ['openai', 'anthropic', 'google', 'mistral', 'cohere','groq', 'perplexity', 'azure', 'together', 'fireworks',- 'deepseek', 'xai', 'huggingface', 'replicate', 'minimax',+ 'deepseek', 'xai', 'huggingface', 'replicate', 'minimax', 'orcarouter',];for (const p of providers) {expect(() => getModelInstance(p, 'key', 'model')).not.toThrow();@@ -120,6 +120,14 @@ describe('getModelInstance', () => {);});+ it('supports OrcaRouter via the OpenAI-compatible endpoint', () => {+ const instance = getModelInstance('orcarouter', 'sk-orca-test', 'orcarouter/auto');+ expect(instance).toBeDefined();+ expect(createOpenAI).toHaveBeenCalledWith(+ expect.objectContaining({ baseURL: 'https://api.orcarouter.ai/v1' })+ );+ });+it('throws when a provider factory returns a non-callable object instance', () => {(createOpenAI as jest.Mock).mockReturnValueOnce({ notAFunction: true });expect(() => getModelInstance('openai', 'key', 'model')).toThrow();diff --git a/src/client/src/__tests__/lib/platform/evaluation/run-evaluation.test.ts b/src/client/src/__tests__/lib/platform/evaluation/run-evaluation.test.tsindex d7f8c0b5..328d29ad 100644--- a/src/client/src/__tests__/lib/platform/evaluation/run-evaluation.test.ts+++ b/src/client/src/__tests__/lib/platform/evaluation/run-evaluation.test.ts@@ -119,6 +119,11 @@ describe('runEvaluation — provider routing', () => {expect(createOpenAI).toHaveBeenCalledWith(expect.objectContaining({ baseURL: 'https://api.fireworks.ai/inference/v1' }));});+ it('creates OpenAI-compatible model for provider=orcarouter', async () => {+ await runEvaluation({ ...BASE_PARAMS, provider: 'orcarouter' });+ expect(createOpenAI).toHaveBeenCalledWith(expect.objectContaining({ baseURL: 'https://api.orcarouter.ai/v1' }));+ });+it('maps gemini provider alias to google', async () => {await runEvaluation({ ...BASE_PARAMS, provider: 'gemini' });expect(createGoogleGenerativeAI).toHaveBeenCalled();diff --git a/src/client/src/__tests__/lib/platform/pricing/chat-cost.test.ts b/src/client/src/__tests__/lib/platform/pricing/chat-cost.test.tsindex e9d00ea1..3fda7ddb 100644--- a/src/client/src/__tests__/lib/platform/pricing/chat-cost.test.ts+++ b/src/client/src/__tests__/lib/platform/pricing/chat-cost.test.ts@@ -73,6 +73,7 @@ describe("promptTokensIncludeCache", () => {expect(promptTokensIncludeCache("openai", 6200, 5000, 1000)).toBe(true);expect(promptTokensIncludeCache("litellm", 6200, 5000, 1000)).toBe(true);expect(promptTokensIncludeCache("minimax", 6200, 5000, 1000)).toBe(true);+ expect(promptTokensIncludeCache("orcarouter", 6200, 5000, 1000)).toBe(true);});it("falls back to heuristic when provider unknown", () => {diff --git a/src/client/src/__tests__/lib/platform/providers/default-models.test.ts b/src/client/src/__tests__/lib/platform/providers/default-models.test.tsindex 30cf6cc3..ff58c225 100644--- a/src/client/src/__tests__/lib/platform/providers/default-models.test.ts+++ b/src/client/src/__tests__/lib/platform/providers/default-models.test.ts@@ -11,7 +11,7 @@ describe('default-models', () => {const expectedProviders = ['openai', 'anthropic', 'google', 'mistral', 'groq','perplexity', 'azure', 'cohere', 'together', 'fireworks',- 'deepseek', 'xai', 'huggingface', 'replicate', 'minimax',+ 'deepseek', 'xai', 'huggingface', 'replicate', 'minimax', 'orcarouter',];for (const provider of expectedProviders) {expect(DEFAULT_MODELS_BY_PROVIDER).toHaveProperty(provider);@@ -91,4 +91,41 @@ describe('default-models', () => {);expect(m27?.capabilities).not.toContain('vision');});++ it('seeds OrcaRouter with documented provider-prefixed models', () => {+ const orcaProvider = DEFAULT_PROVIDERS.find((p) => p.providerId === 'orcarouter');+ expect(orcaProvider).toEqual(+ expect.objectContaining({+ displayName: 'OrcaRouter',+ requiresVault: true,+ })+ );++ const orcaModels = DEFAULT_MODELS_BY_PROVIDER.orcarouter;+ expect(orcaModels.map((m) => m.id)).toEqual([+ 'orcarouter/auto',+ 'openai/gpt-5.5',+ 'anthropic/claude-opus-4.7',+ 'google/gemini-2.5-flash',+ ]);++ expect(orcaModels.find((m) => m.id === 'openai/gpt-5.5')).toMatchObject({+ contextWindow: 400000,+ inputPricePerMToken: 5.0,+ outputPricePerMToken: 30.0,+ cacheReadPricePerMToken: 0.5,+ });+ expect(orcaModels.find((m) => m.id === 'anthropic/claude-opus-4.7')).toMatchObject({+ contextWindow: 1000000,+ inputPricePerMToken: 5.0,+ outputPricePerMToken: 25.0,+ cacheReadPricePerMToken: 0.5,+ cacheCreationPricePerMToken: 6.25,+ });+ expect(orcaModels.find((m) => m.id === 'google/gemini-2.5-flash')).toMatchObject({+ contextWindow: 1000000,+ inputPricePerMToken: 0.3,+ outputPricePerMToken: 2.5,+ });+ });});diff --git a/src/client/src/app/(playground)/agents/controller/[instance_id]/page.tsx b/src/client/src/app/(playground)/agents/controller/[instance_id]/page.tsxindex 6d3c8fcc..7c3eb12d 100644--- a/src/client/src/app/(playground)/agents/controller/[instance_id]/page.tsx+++ b/src/client/src/app/(playground)/agents/controller/[instance_id]/page.tsx@@ -58,6 +58,7 @@ const OPENAI_COMPATIBLE_PROVIDERS = new Set(["deepseek","together","fireworks",+ "orcarouter","azure_inference","azure_openai","vercel_ai",@@ -94,6 +95,7 @@ const PROVIDER_LABELS: Record<string, string> = {deepseek: getMessage().AGENTS_PROVIDER_DEEPSEEK,together: getMessage().AGENTS_PROVIDER_TOGETHER,fireworks: getMessage().AGENTS_PROVIDER_FIREWORKS,+ orcarouter: getMessage().AGENTS_PROVIDER_ORCAROUTER,vercel_ai: getMessage().AGENTS_PROVIDER_VERCEL_AI,vertex_ai: getMessage().AGENTS_PROVIDER_VERTEX_AI,azure_inference: getMessage().AGENTS_PROVIDER_AZURE_INFERENCE,diff --git a/src/client/src/clickhouse/migrations/index.ts b/src/client/src/clickhouse/migrations/index.tsindex d47f3497..efe55dc5 100644--- a/src/client/src/clickhouse/migrations/index.ts+++ b/src/client/src/clickhouse/migrations/index.ts@@ -34,6 +34,7 @@ import AddCodingAgentLOCSummaryFieldsMigration from "./add-coding-agent-loc-summimport CreateCodingAgentsAuditMigration from "./create-coding-agents-audit-migration";import DropVcsMigration from "./drop-vcs-migration";import AddProviderModelsCachePricesMigration from "./add-provider-models-cache-prices-migration";+import SeedOrcaRouterProviderMigration from "./seed-orcarouter-provider-migration";import CreateTelemetryRollupsMigration from "./create-telemetry-rollups-migration";import AlterTelemetryRollupsDimensionsMigration from "./alter-telemetry-rollups-dimensions-migration";@@ -73,6 +74,7 @@ export default async function migrations(databaseConfigId?: string) {CreateProviderMetadataMigration(databaseConfigId),DropLegacyOpengroundTablesMigration(databaseConfigId),]);+ await SeedOrcaRouterProviderMigration(databaseConfigId);await EncryptVaultValuesMigration(databaseConfigId);await AddChatConversationTypeMigration(databaseConfigId);diff --git a/src/client/src/clickhouse/migrations/seed-orcarouter-provider-migration.ts b/src/client/src/clickhouse/migrations/seed-orcarouter-provider-migration.tsnew file mode 100644index 00000000..c7d1f4a2--- /dev/null+++ b/src/client/src/clickhouse/migrations/seed-orcarouter-provider-migration.ts@@ -0,0 +1,156 @@+import getMessage from "@/constants/messages";+import { getDBConfigByIdInternal, getDBConfigByUser } from "@/lib/db-config";+import { dataCollector } from "@/lib/platform/common";+import {+ DEFAULT_MODELS_BY_PROVIDER,+ DEFAULT_PROVIDERS,+} from "@/lib/platform/providers/default-models";+import {+ OPENLIT_PROVIDER_METADATA_TABLE_NAME,+ OPENLIT_PROVIDER_MODELS_TABLE_NAME,+} from "@/lib/platform/providers/table-details";+import prisma from "@/lib/prisma";+import asaw from "@/utils/asaw";+import { consoleLog } from "@/utils/log";++const MIGRATION_ID = "seed-orcarouter-provider";+const PROVIDER_ID = "orcarouter";++/**+ * Existing ClickHouse instances already ran the one-shot provider-metadata+ * and provider-models seeds, so they never pick up providers added to the+ * static registries later. Insert OrcaRouter (and only OrcaRouter) when it+ * is missing, without overwriting user-edited rows.+ */+export default async function SeedOrcaRouterProviderMigration(+ databaseConfigId?: string+) {+ const [, dbConfig] = await asaw(+ databaseConfigId+ ? getDBConfigByIdInternal({ id: databaseConfigId })+ : getDBConfigByUser(true)+ );++ if (!dbConfig?.id) return { err: getMessage().DATABASE_CONFIG_NOT_FOUND };++ const [, migrationExist] = await asaw(+ prisma.clickhouseMigrations.findFirst({+ where: {+ AND: {+ databaseConfigId: dbConfig.id,+ clickhouseMigrationId: MIGRATION_ID,+ },+ },+ })+ );++ if (migrationExist?.id) {+ return { migrationExist: true };+ }++ const provider = DEFAULT_PROVIDERS.find((p) => p.providerId === PROVIDER_ID);+ const models = DEFAULT_MODELS_BY_PROVIDER[PROVIDER_ID] || [];+ if (!provider) {+ return { err: getMessage().OPERATION_FAILED };+ }++ const { data: existingProviders, err: providerQueryErr } = await dataCollector(+ {+ query: `+ SELECT provider_id+ FROM ${OPENLIT_PROVIDER_METADATA_TABLE_NAME}+ WHERE provider_id = '${PROVIDER_ID}'+ `,+ },+ "query",+ dbConfig.id+ );+ if (providerQueryErr) {+ consoleLog(providerQueryErr);+ return { err: getMessage().OPERATION_FAILED };+ }++ if (!((existingProviders as { provider_id: string }[]) || []).length) {+ const { err: insertProviderErr } = await dataCollector(+ {+ table: OPENLIT_PROVIDER_METADATA_TABLE_NAME,+ values: [+ {+ provider_id: provider.providerId,+ display_name: provider.displayName,+ description: provider.description,+ requires_vault: provider.requiresVault,+ config_schema: JSON.stringify(provider.configSchema),+ is_default: true,+ },+ ],+ },+ "insert",+ dbConfig.id+ );+ if (insertProviderErr) {+ console.error("Error seeding OrcaRouter provider metadata:", insertProviderErr);+ return { err: getMessage().OPERATION_FAILED };+ }+ }++ const { data: existingModels, err: modelQueryErr } = await dataCollector(+ {+ query: `+ SELECT model_id+ FROM ${OPENLIT_PROVIDER_MODELS_TABLE_NAME}+ WHERE provider = '${PROVIDER_ID}'+ `,+ },+ "query",+ dbConfig.id+ );+ if (modelQueryErr) {+ consoleLog(modelQueryErr);+ return { err: getMessage().OPERATION_FAILED };+ }++ const existingModelIds = new Set(+ ((existingModels as { model_id: string }[]) || []).map((row) => row.model_id)+ );+ const toSeed = models+ .filter((model) => !existingModelIds.has(model.id))+ .map((model) => ({+ provider: PROVIDER_ID,+ model_id: model.id,+ display_name: model.displayName,+ model_type: "chat",+ context_window: model.contextWindow,+ input_price_per_m_token: model.inputPricePerMToken,+ output_price_per_m_token: model.outputPricePerMToken,+ cache_read_price_per_m_token: model.cacheReadPricePerMToken || 0,+ cache_creation_price_per_m_token: model.cacheCreationPricePerMToken || 0,+ capabilities: model.capabilities || [],+ is_default: true,+ created_by_user_id: "",+ }));++ if (toSeed.length > 0) {+ const { err: seedErr } = await dataCollector(+ {+ table: OPENLIT_PROVIDER_MODELS_TABLE_NAME,+ values: toSeed,+ },+ "insert",+ dbConfig.id+ );+ if (seedErr) {+ console.error("Error seeding OrcaRouter models:", seedErr);+ return { err: getMessage().OPERATION_FAILED };+ }+ }++ await prisma.clickhouseMigrations.create({+ data: {+ databaseConfigId: dbConfig.id,+ clickhouseMigrationId: MIGRATION_ID,+ },+ });++ return { data: "OrcaRouter provider seed successful" };+}diff --git a/src/client/src/components/svg/providers.tsx b/src/client/src/components/svg/providers.tsxindex 7baf186e..cccef287 100644--- a/src/client/src/components/svg/providers.tsx+++ b/src/client/src/components/svg/providers.tsx@@ -387,6 +387,31 @@ export function CustomGatewaySvg({ className }: SvgProps) {);}+export function OrcaRouterSvg({ className }: SvgProps) {+ // Theme-safe mark (currentColor) so it reads on light and dark surfaces.+ // Distinct from CustomGatewaySvg: a single routed body rather than a hub.+ return (+ <svg+ viewBox="0 0 24 24"+ xmlns="http://www.w3.org/2000/svg"+ className={className}+ fill="currentColor"+ >+ <path d="M3.2 13.2c2.2-3.4 5.4-5.1 9.3-5.1 3.1 0 5.8 1 7.6 2.7-.4 2.1-2.6 3.8-5.6 4.6-2 .5-4.2.5-6.2 0-2.1-.6-3.8-1.2-5.1-2.2z" />+ <path d="M11.4 8.2c.3-1.6 1-3 2.1-4.1.2 1.3.1 2.7-.2 4.1" />+ <circle cx="8.4" cy="11.1" r="0.85" />+ <path+ d="M16.2 10.4c1.6.1 2.9.6 3.5 1.4"+ fill="none"+ stroke="currentColor"+ strokeWidth="1.4"+ strokeLinecap="round"+ strokeDasharray="2 1.6"+ />+ </svg>+ );+}+const PROVIDER_ICON_MAP: Record<string, React.FC<SvgProps>> = {openai: OpenAISvg,anthropic: AnthropicSvg,@@ -394,6 +419,7 @@ const PROVIDER_ICON_MAP: Record<string, React.FC<SvgProps>> = {cohere: CohereSvg,mistral: MistralSvg,groq: GroqSvg,+ orcarouter: OrcaRouterSvg,azure_openai: AzureSvg,azure_inference: AzureSvg,bedrock: BedrockSvg,diff --git a/src/client/src/constants/messages/en.ts b/src/client/src/constants/messages/en.tsindex 368eed19..0acf5a84 100644--- a/src/client/src/constants/messages/en.ts+++ b/src/client/src/constants/messages/en.ts@@ -1134,6 +1134,7 @@ export const AGENTS_PROVIDER_GROQ = "Groq";export const AGENTS_PROVIDER_DEEPSEEK = "DeepSeek";export const AGENTS_PROVIDER_TOGETHER = "Together AI";export const AGENTS_PROVIDER_FIREWORKS = "Fireworks AI";+export const AGENTS_PROVIDER_ORCAROUTER = "OrcaRouter";export const AGENTS_PROVIDER_VERCEL_AI = "Vercel AI Gateway";export const AGENTS_PROVIDER_VERTEX_AI = "Vertex AI";export const AGENTS_PROVIDER_AZURE_INFERENCE = "Azure AI Inference";diff --git a/src/client/src/lib/platform/chat/stream.ts b/src/client/src/lib/platform/chat/stream.tsindex 4905899f..f3c49687 100644--- a/src/client/src/lib/platform/chat/stream.ts+++ b/src/client/src/lib/platform/chat/stream.ts@@ -40,6 +40,7 @@ const providerFactories: Record<string, ProviderFactory> = {huggingface: (apiKey) => createOpenAI({ baseURL: "https://api-inference.huggingface.co/v1", apiKey }),replicate: (apiKey) => createOpenAI({ baseURL: "https://openai-proxy.replicate.com/v1", apiKey }),minimax: (apiKey) => createOpenAI({ baseURL: "https://api.minimax.io/v1", apiKey }),+ orcarouter: (apiKey) => createOpenAI({ baseURL: "https://api.orcarouter.ai/v1", apiKey }),};const VALID_PROVIDERS = Object.keys(providerFactories);diff --git a/src/client/src/lib/platform/evaluation/run-evaluation.ts b/src/client/src/lib/platform/evaluation/run-evaluation.tsindex c929ce90..04e573f0 100644--- a/src/client/src/lib/platform/evaluation/run-evaluation.ts+++ b/src/client/src/lib/platform/evaluation/run-evaluation.ts@@ -108,6 +108,11 @@ function getModel(provider: string, model: string, apiKey: string) {baseURL: "https://api.fireworks.ai/inference/v1",apiKey,})(model);+ case "orcarouter":+ return createOpenAI({+ baseURL: "https://api.orcarouter.ai/v1",+ apiKey,+ })(model);default:throw new Error(`Provider ${provider} not supported for evaluation`);}diff --git a/src/client/src/lib/platform/openground/ai-sdk-adapter.ts b/src/client/src/lib/platform/openground/ai-sdk-adapter.tsindex 64498200..afd9a52a 100644--- a/src/client/src/lib/platform/openground/ai-sdk-adapter.ts+++ b/src/client/src/lib/platform/openground/ai-sdk-adapter.ts@@ -78,12 +78,17 @@ export class AISdkAdapter {baseURL: 'https://api.minimax.io/v1',apiKey}),+ orcarouter: (apiKey: string) => createOpenAI({+ baseURL: 'https://api.orcarouter.ai/v1',+ apiKey+ }),};private static readonly SUPPORTED_PROVIDERS = ['openai', 'anthropic', 'google', 'mistral', 'cohere','groq', 'perplexity', 'azure', 'together', 'fireworks','deepseek', 'xai', 'huggingface', 'replicate', 'minimax',+ 'orcarouter',];/**diff --git a/src/client/src/lib/platform/pricing/chat-cost.ts b/src/client/src/lib/platform/pricing/chat-cost.tsindex 700a5890..15cfe3e2 100644--- a/src/client/src/lib/platform/pricing/chat-cost.ts+++ b/src/client/src/lib/platform/pricing/chat-cost.ts@@ -44,6 +44,7 @@ const INCLUSIVE_CACHE_PROVIDERS = new Set(["cohere","xai","minimax",+ "orcarouter","claude_agent_sdk","claude-agent-sdk","pydo",diff --git a/src/client/src/lib/platform/providers/default-models.ts b/src/client/src/lib/platform/providers/default-models.tsindex d6da306d..9cd6c61a 100644--- a/src/client/src/lib/platform/providers/default-models.ts+++ b/src/client/src/lib/platform/providers/default-models.ts@@ -143,6 +143,14 @@ export const DEFAULT_PROVIDERS: DefaultProviderEntry[] = [topP: { min: 0, max: 1, step: 0.1, default: 0.95, description: "Nucleus sampling threshold" },},},+ {+ providerId: "orcarouter", displayName: "OrcaRouter", description: "OpenAI-compatible AI gateway with adaptive routing across 200+ models", requiresVault: true,+ configSchema: {+ temperature: { min: 0, max: 2, step: 0.1, default: 1, description: "Sampling temperature" },+ maxTokens: { min: 1, max: 32768, step: 1, default: 1000, description: "Maximum tokens to generate" },+ topP: { min: 0, max: 1, step: 0.1, default: 1, description: "Nucleus sampling threshold" },+ },+ },];export const DEFAULT_MODELS_BY_PROVIDER: Record<string, DefaultModelEntry[]> = {@@ -1103,4 +1111,50 @@ export const DEFAULT_MODELS_BY_PROVIDER: Record<string, DefaultModelEntry[]> = {capabilities: ["function-calling", "streaming", "thinking"],},],+ // OrcaRouter model IDs are provider-prefixed (`openai/…`, `anthropic/…`)+ // per https://docs.orcarouter.ai/getting-started/models. Prices match the+ // upstream OpenLIT catalog because OrcaRouter bills at zero markup.+ orcarouter: [+ {+ id: "orcarouter/auto",+ displayName: "Auto Router",+ contextWindow: 200000,+ // Virtual router: billed cost follows the resolved upstream+ // (`X-Orca-Resolved-Model`). Seed gpt-4o-mini rates so the+ // picker has a non-zero estimate rather than $0.+ inputPricePerMToken: 0.15,+ outputPricePerMToken: 0.6,+ cacheReadPricePerMToken: 0.075,+ capabilities: ["streaming"],+ },+ {+ id: "openai/gpt-5.5",+ displayName: "GPT-5.5",+ contextWindow: 400000,+ inputPricePerMToken: 5.0,+ outputPricePerMToken: 30.0,+ cacheReadPricePerMToken: 0.5,+ capabilities: ["function-calling", "vision", "streaming"],+ },+ {+ id: "anthropic/claude-opus-4.7",+ displayName: "Claude Opus 4.7",+ contextWindow: 1000000,+ inputPricePerMToken: 5.0,+ outputPricePerMToken: 25.0,+ cacheReadPricePerMToken: 0.5,+ cacheCreationPricePerMToken: 6.25,+ capabilities: ["function-calling", "vision", "streaming", "thinking"],+ },+ {+ id: "google/gemini-2.5-flash",+ displayName: "Gemini 2.5 Flash",+ contextWindow: 1000000,+ inputPricePerMToken: 0.3,+ outputPricePerMToken: 2.5,+ cacheReadPricePerMToken: 0.03,+ cacheCreationPricePerMToken: 0.3,+ capabilities: ["function-calling", "vision", "streaming", "thinking"],+ },+ ],};