diff --git a/src/agent/core/domain/llm/registry.ts b/src/agent/core/domain/llm/registry.ts
index bc37b4083..cd4b8e447 100644
--- a/src/agent/core/domain/llm/registry.ts
+++ b/src/agent/core/domain/llm/registry.ts
@@ -735,6 +735,7 @@ export function resolveRegistryProvider(
         'openai-compatible',
         'openrouter',
         'perplexity',
+        'requesty',
         'togetherai',
         'vercel',
         'xai',
diff --git a/src/agent/infra/llm/providers/index.ts b/src/agent/infra/llm/providers/index.ts
index d6cdef877..682644cd6 100644
--- a/src/agent/infra/llm/providers/index.ts
+++ b/src/agent/infra/llm/providers/index.ts
@@ -26,6 +26,7 @@ import {openaiCompatibleProvider} from './openai-compatible.js'
 import {openaiProvider} from './openai.js'
 import {openrouterProvider} from './openrouter.js'
 import {perplexityProvider} from './perplexity.js'
+import {requestyProvider} from './requesty.js'
 import {togetheraiProvider} from './togetherai.js'
 import {vercelProvider} from './vercel.js'
 import {xaiProvider} from './xai.js'
@@ -52,6 +53,7 @@ const PROVIDER_MODULES: Readonly<Record<string, ProviderModule>> = {
   'openai-compatible': openaiCompatibleProvider,
   openrouter: openrouterProvider,
   perplexity: perplexityProvider,
+  requesty: requestyProvider,
   togetherai: togetheraiProvider,
   vercel: vercelProvider,
   xai: xaiProvider,
diff --git a/src/agent/infra/llm/providers/requesty.ts b/src/agent/infra/llm/providers/requesty.ts
new file mode 100644
index 000000000..45dd64eea
--- /dev/null
+++ b/src/agent/infra/llm/providers/requesty.ts
@@ -0,0 +1,47 @@
+/**
+ * Requesty Provider Module
+ *
+ * Access 300+ models via the Requesty router using its OpenAI-compatible API.
+ * Requesty exposes an OpenAI-shaped endpoint at https://router.requesty.ai/v1,
+ * so it is wired through @ai-sdk/openai-compatible.
+ */
+
+import {createOpenAICompatible} from '@ai-sdk/openai-compatible'
+
+import type {GeneratorFactoryConfig, ProviderModule} from './types.js'
+
+import {AiSdkContentGenerator} from '../generators/ai-sdk-content-generator.js'
+
+const REQUESTY_BASE_URL = 'https://router.requesty.ai/v1'
+
+export const requestyProvider: ProviderModule = {
+  apiKeyUrl: 'https://app.requesty.ai/api-keys',
+  authType: 'api-key',
+  baseUrl: REQUESTY_BASE_URL,
+  category: 'popular',
+  createGenerator(config: GeneratorFactoryConfig) {
+    const provider = createOpenAICompatible({
+      apiKey: config.apiKey || '',
+      baseURL: config.baseUrl || REQUESTY_BASE_URL,
+      headers: {
+        'HTTP-Referer': config.httpReferer ?? 'https://byterover.dev',
+        'X-Title': config.siteName ?? 'byterover-cli',
+        ...config.headers,
+      },
+      name: 'requesty',
+    })
+
+    return new AiSdkContentGenerator({
+      model: provider.chatModel(config.model),
+      requestTimeoutMs: config.requestTimeoutMs,
+    })
+  },
+  defaultModel: 'openai/gpt-4o-mini',
+  description: 'Access 300+ models via the Requesty router',
+  envVars: ['REQUESTY_API_KEY'],
+  id: 'requesty',
+  name: 'Requesty',
+  priority: 1.5,
+
+  providerType: 'openai',
+}
diff --git a/src/server/core/domain/entities/provider-registry.ts b/src/server/core/domain/entities/provider-registry.ts
index aff9fe1d6..bbd92f956 100644
--- a/src/server/core/domain/entities/provider-registry.ts
+++ b/src/server/core/domain/entities/provider-registry.ts
@@ -321,6 +321,22 @@ export const PROVIDER_REGISTRY: Readonly<Record<string, ProviderDefinition>> = {
     name: 'Perplexity',
     priority: 13,
   },
+  requesty: {
+    apiKeyUrl: 'https://app.requesty.ai/api-keys',
+    baseUrl: 'https://router.requesty.ai/v1',
+    category: 'popular',
+    defaultModel: 'openai/gpt-4o-mini',
+    description: 'Access 300+ models via the Requesty router',
+    envVars: ['REQUESTY_API_KEY'],
+    headers: {
+      'HTTP-Referer': 'https://byterover.dev',
+      'X-Title': 'byterover-cli',
+    },
+    id: 'requesty',
+    modelsEndpoint: '/models',
+    name: 'Requesty',
+    priority: 1.5,
+  },
   togetherai: {
     apiKeyUrl: 'https://api.together.ai/settings/api-keys',
     baseUrl: 'https://api.together.xyz/v1',
diff --git a/src/server/infra/http/provider-model-fetcher-registry.ts b/src/server/infra/http/provider-model-fetcher-registry.ts
index 027e0aee9..c13516b47 100644
--- a/src/server/infra/http/provider-model-fetcher-registry.ts
+++ b/src/server/infra/http/provider-model-fetcher-registry.ts
@@ -16,6 +16,7 @@ import {
   OpenAICompatibleModelFetcher,
   OpenAIModelFetcher,
   OpenRouterModelFetcher,
+  RequestyModelFetcher,
 } from './provider-model-fetchers.js'
 
 /**
@@ -134,6 +135,12 @@ export async function getModelFetcher(providerId: string): Promise<IProviderMode
       break
     }
 
+    case 'requesty': {
+      fetcher = new RequestyModelFetcher()
+
+      break
+    }
+
     case 'perplexity': {
       fetcher = new ChatBasedModelFetcher(
         'https://api.perplexity.ai',
diff --git a/src/server/infra/http/provider-model-fetchers.ts b/src/server/infra/http/provider-model-fetchers.ts
index e62651c50..89dddfda7 100644
--- a/src/server/infra/http/provider-model-fetchers.ts
+++ b/src/server/infra/http/provider-model-fetchers.ts
@@ -610,6 +610,37 @@ export class OpenRouterModelFetcher implements IProviderModelFetcher {
   }
 }
 
+// ============================================================================
+// Requesty Model Fetcher (wraps existing client)
+// ============================================================================
+
+import {getRequestyApiClient, type NormalizedModel as RequestyNormalizedModel} from './requesty-api-client.js'
+
+/**
+ * Model fetcher that wraps the existing RequestyApiClient.
+ * Adapts NormalizedModel to ProviderModelInfo.
+ */
+export class RequestyModelFetcher implements IProviderModelFetcher {
+  async fetchModels(apiKey: string, options?: FetchModelsOptions): Promise<ProviderModelInfo[]> {
+    const client = getRequestyApiClient()
+    const models = await client.fetchModels(apiKey, options?.forceRefresh ?? false)
+    return models.map((m: RequestyNormalizedModel) => ({
+      contextLength: m.contextLength,
+      description: m.description,
+      id: m.id,
+      isFree: m.isFree,
+      name: m.name,
+      pricing: m.pricing,
+      provider: m.provider,
+    }))
+  }
+
+  async validateApiKey(apiKey: string): Promise<{error?: string; isValid: boolean}> {
+    const client = getRequestyApiClient()
+    return client.validateApiKey(apiKey)
+  }
+}
+
 // ============================================================================
 // Shared helpers
 // ============================================================================
diff --git a/src/server/infra/http/requesty-api-client.ts b/src/server/infra/http/requesty-api-client.ts
new file mode 100644
index 000000000..e6af2cdbf
--- /dev/null
+++ b/src/server/infra/http/requesty-api-client.ts
@@ -0,0 +1,263 @@
+/**
+ * Requesty API Client
+ *
+ * Handles API calls to Requesty for:
+ * - Fetching available models
+ * - Validating API keys
+ *
+ * Uses the Requesty REST API (OpenAI-compatible): https://router.requesty.ai/v1
+ */
+
+import axios, {isAxiosError} from 'axios'
+
+import {ProxyConfig} from './proxy-config.js'
+
+/**
+ * Requesty model from the /models endpoint.
+ * Requesty is OpenAI-compatible and exposes models in the OpenAI shape.
+ */
+export interface RequestyModel {
+  /** Supported modalities */
+  architecture?: {
+    instruct_type?: string
+    modality: string // e.g., 'text->text', 'text+image->text'
+    tokenizer: string
+  }
+  /** Context length (max tokens) */
+  context_length: number
+  /** Description */
+  description?: string
+  /** Model ID (e.g., 'openai/gpt-4o-mini') */
+  id: string
+  /** Display name */
+  name: string
+  /** Per-request limits */
+  per_request_limits?: {
+    completion_tokens?: string
+    prompt_tokens?: string
+  }
+  /** Pricing per token (as string) */
+  pricing: {
+    completion: string // USD per output token (as string)
+    prompt: string // USD per input token (as string)
+  }
+  /** Top provider info */
+  top_provider?: {
+    context_length?: number
+    is_moderated?: boolean
+    max_completion_tokens?: number
+  }
+}
+
+/**
+ * Response from Requesty /models endpoint.
+ */
+interface ModelsResponse {
+  data: RequestyModel[]
+}
+
+/**
+ * Normalized model for use in the application.
+ */
+export interface NormalizedModel {
+  /** Context window size */
+  contextLength: number
+  /** Optional description */
+  description?: string
+  /** Model ID (e.g., 'openai/gpt-4o-mini') */
+  id: string
+  /** Whether this model is free */
+  isFree: boolean
+  /** Display name */
+  name: string
+  /** Pricing per million tokens */
+  pricing: {
+    inputPerM: number
+    outputPerM: number
+  }
+  /** Provider name extracted from ID (e.g., 'openai') */
+  provider: string
+}
+
+/**
+ * Cache entry for models.
+ */
+interface ModelCache {
+  models: NormalizedModel[]
+  timestamp: number
+}
+
+/**
+ * Requesty API client configuration.
+ */
+export interface RequestyApiClientConfig {
+  /** Base URL for Requesty API */
+  baseUrl?: string
+  /** Cache TTL in milliseconds (default: 1 hour) */
+  cacheTtlMs?: number
+  /** HTTP Referer header */
+  httpReferer?: string
+  /** X-Title header */
+  xTitle?: string
+}
+
+const DEFAULT_BASE_URL = 'https://router.requesty.ai/v1'
+const DEFAULT_CACHE_TTL = 60 * 60 * 1000 // 1 hour
+
+/**
+ * Requesty API Client.
+ *
+ * Provides methods to interact with the Requesty API for fetching models
+ * and validating API keys.
+ *
+ * @example
+ * ```typescript
+ * const client = new RequestyApiClient()
+ *
+ * // Validate API key
+ * const isValid = await client.validateApiKey('...')
+ *
+ * // Fetch models
+ * const models = await client.fetchModels('...')
+ * ```
+ */
+export class RequestyApiClient {
+  private readonly baseUrl: string
+  private readonly cacheTtlMs: number
+  private readonly httpReferer?: string
+  private modelCache: ModelCache | undefined
+  private readonly xTitle?: string
+
+  public constructor(config: RequestyApiClientConfig = {}) {
+    this.baseUrl = config.baseUrl ?? DEFAULT_BASE_URL
+    this.cacheTtlMs = config.cacheTtlMs ?? DEFAULT_CACHE_TTL
+    this.httpReferer = config.httpReferer ?? 'https://byterover.dev'
+    this.xTitle = config.xTitle ?? 'byterover-cli'
+  }
+
+  /**
+   * Clears the model cache.
+   */
+  public clearCache(): void {
+    this.modelCache = undefined
+  }
+
+  /**
+   * Fetches available models from Requesty.
+   * Results are cached for the configured TTL.
+   *
+   * @param apiKey - The API key to use
+   * @param forceRefresh - If true, bypasses cache
+   * @returns Array of normalized models
+   */
+  public async fetchModels(apiKey: string, forceRefresh = false): Promise<NormalizedModel[]> {
+    // Check cache
+    if (!forceRefresh && this.modelCache && Date.now() - this.modelCache.timestamp < this.cacheTtlMs) {
+      return this.modelCache.models
+    }
+
+    const models = await this.fetchModelsInternal(apiKey)
+
+    // Update cache
+    this.modelCache = {
+      models,
+      timestamp: Date.now(),
+    }
+
+    return models
+  }
+
+  /**
+   * Validates an API key by attempting to fetch models.
+   *
+   * @param apiKey - The API key to validate
+   * @returns Object with isValid flag and optional error message
+   */
+  public async validateApiKey(apiKey: string): Promise<{error?: string; isValid: boolean}> {
+    try {
+      await this.fetchModelsInternal(apiKey)
+      return {isValid: true}
+    } catch (error) {
+      if (isAxiosError(error)) {
+        if (error.response?.status === 401) {
+          return {error: 'Invalid API key', isValid: false}
+        }
+
+        if (error.response?.status === 403) {
+          return {error: 'API key does not have required permissions', isValid: false}
+        }
+
+        return {error: `API error: ${error.response?.statusText ?? error.message}`, isValid: false}
+      }
+
+      return {error: error instanceof Error ? error.message : 'Unknown error', isValid: false}
+    }
+  }
+
+  /**
+   * Internal method to fetch models from Requesty API.
+   */
+  private async fetchModelsInternal(apiKey: string): Promise<NormalizedModel[]> {
+    const response = await axios.get<ModelsResponse>(`${this.baseUrl}/models`, {
+      headers: {
+        Authorization: `Bearer ${apiKey}`,
+        'HTTP-Referer': this.httpReferer,
+        'X-Title': this.xTitle,
+      },
+      httpAgent: ProxyConfig.getProxyAgent(),
+      httpsAgent: ProxyConfig.getProxyAgent(),
+      proxy: false,
+      timeout: 30_000,
+    })
+
+    return response.data.data.map((model) => this.normalizeModel(model))
+  }
+
+  /**
+   * Normalizes a Requesty model to our standard format.
+   */
+  private normalizeModel(model: RequestyModel): NormalizedModel {
+    // Extract provider from model ID (e.g., 'openai' from 'openai/gpt-4o-mini')
+    const [provider, ...nameParts] = model.id.split('/')
+    const shortName = nameParts.join('/') || model.id
+
+    // Parse pricing (convert from string to number)
+    // Requesty returns price per token, multiply by 1M to get price per million tokens
+    const inputPricePerToken = Number.parseFloat(model.pricing.prompt) || 0
+    const outputPricePerToken = Number.parseFloat(model.pricing.completion) || 0
+    const inputPerM = inputPricePerToken * 1_000_000
+    const outputPerM = outputPricePerToken * 1_000_000
+
+    // Check if free (both prices are 0)
+    const isFree = inputPricePerToken === 0 && outputPricePerToken === 0
+
+    return {
+      contextLength: model.context_length,
+      description: model.description,
+      id: model.id,
+      isFree,
+      name: model.name || shortName,
+      pricing: {
+        inputPerM,
+        outputPerM,
+      },
+      provider: `Requesty (${provider})`,
+    }
+  }
+}
+
+/**
+ * Singleton instance of the Requesty API client.
+ */
+let _requestyApiClient: RequestyApiClient | undefined
+
+/**
+ * Gets or creates the singleton Requesty API client.
+ */
+export function getRequestyApiClient(): RequestyApiClient {
+  if (!_requestyApiClient) {
+    _requestyApiClient = new RequestyApiClient()
+  }
+
+  return _requestyApiClient
+}
diff --git a/src/server/infra/provider/provider-config-resolver.ts b/src/server/infra/provider/provider-config-resolver.ts
index 6790b00b9..7792324ba 100644
--- a/src/server/infra/provider/provider-config-resolver.ts
+++ b/src/server/infra/provider/provider-config-resolver.ts
@@ -154,6 +154,21 @@ export async function resolveProviderConfig(
       }
     }
 
+    case 'requesty': {
+      const providerDef = getProviderById(activeProvider)
+      const headers = providerDef?.headers
+      return {
+        activeModel,
+        activeProvider,
+        maxInputTokens,
+        provider: activeProvider,
+        providerApiKey: apiKey || undefined,
+        providerBaseUrl: config.getBaseUrl(activeProvider) || providerDef?.baseUrl || undefined,
+        providerHeaders: headers && Object.keys(headers).length > 0 ? {...headers} : undefined,
+        providerKeyMissing: providerRequiresApiKey(activeProvider) && !apiKey,
+      }
+    }
+
     default: {
       const providerDef = getProviderById(activeProvider)
       const providerConfig = config.providers[activeProvider]