diff --git a/packages/rest/src/__tests__/client.spec.ts b/packages/rest/src/__tests__/client.spec.ts
index babb3ea..5e00d27 100644
--- a/packages/rest/src/__tests__/client.spec.ts
+++ b/packages/rest/src/__tests__/client.spec.ts
@@ -305,13 +305,16 @@ describe('@figmarine/rest - client', () => {
   });
 
   describe('Options - rateLimit', () => {
-    it('rate limits by default', async ({ mockedEnv }) => {
+    // Note that we cannot easily spy on axiosRetry because it's a default export.
+    it('only applies 429 exponential retry by default', async ({ mockedEnv }) => {
       const logSpy = vi.spyOn(loggerModule, 'log');
       const rlSpy = vi.spyOn(interceptorsModule, 'rateLimitRequestInterceptor');
       mockedEnv({});
       await Client();
-      expect(logSpy).toHaveBeenCalledWith('Applying rate limit proxy to API client.');
-      expect(rlSpy).toHaveBeenCalled();
+      expect(logSpy).toHaveBeenCalledWith('Applying rate limit safeguards to API client.');
+      expect(logSpy).toHaveBeenCalledWith('Applying exponential retry on Error 429.');
+      expect(logSpy).not.toHaveBeenCalledWith('Applying proactive rate limit (limiting req/s).');
+      expect(rlSpy).not.toHaveBeenCalled();
     });
 
     it('does not rate limit when false', async ({ mockedEnv }) => {
@@ -319,16 +322,42 @@ describe('@figmarine/rest - client', () => {
       const rlSpy = vi.spyOn(interceptorsModule, 'rateLimitRequestInterceptor');
       mockedEnv({});
       await Client({ rateLimit: false });
-      expect(logSpy).not.toHaveBeenCalledWith('Applying rate limit proxy to API client.');
+      expect(logSpy).not.toHaveBeenCalledWith('Applying rate limit safeguards to API client.');
+      expect(logSpy).not.toHaveBeenCalledWith('Applying exponential retry on Error 429.');
+      expect(logSpy).not.toHaveBeenCalledWith('Applying proactive rate limit (limiting req/s).');
       expect(rlSpy).not.toHaveBeenCalled();
     });
 
-    it('does rate limit when true', async ({ mockedEnv }) => {
+    it('does rate limit with all options when true', async ({ mockedEnv }) => {
       const logSpy = vi.spyOn(loggerModule, 'log');
       const rlSpy = vi.spyOn(interceptorsModule, 'rateLimitRequestInterceptor');
       mockedEnv({});
       await Client({ rateLimit: true });
-      expect(logSpy).toHaveBeenCalledWith('Applying rate limit proxy to API client.');
+      expect(logSpy).toHaveBeenCalledWith('Applying rate limit safeguards to API client.');
+      expect(logSpy).toHaveBeenCalledWith('Applying exponential retry on Error 429.');
+      expect(logSpy).toHaveBeenCalledWith('Applying proactive rate limit (limiting req/s).');
+      expect(rlSpy).toHaveBeenCalled();
+    });
+
+    it('only applies 429 exponential retry when set to "reactive"', async ({ mockedEnv }) => {
+      const logSpy = vi.spyOn(loggerModule, 'log');
+      const rlSpy = vi.spyOn(interceptorsModule, 'rateLimitRequestInterceptor');
+      mockedEnv({});
+      await Client({ rateLimit: 'reactive' });
+      expect(logSpy).toHaveBeenCalledWith('Applying rate limit safeguards to API client.');
+      expect(logSpy).toHaveBeenCalledWith('Applying exponential retry on Error 429.');
+      expect(logSpy).not.toHaveBeenCalledWith('Applying proactive rate limit (limiting req/s).');
+      expect(rlSpy).not.toHaveBeenCalled();
+    });
+
+    it('does rate limit with all options when set to "proactive"', async ({ mockedEnv }) => {
+      const logSpy = vi.spyOn(loggerModule, 'log');
+      const rlSpy = vi.spyOn(interceptorsModule, 'rateLimitRequestInterceptor');
+      mockedEnv({});
+      await Client({ rateLimit: 'proactive' });
+      expect(logSpy).toHaveBeenCalledWith('Applying rate limit safeguards to API client.');
+      expect(logSpy).toHaveBeenCalledWith('Applying exponential retry on Error 429.');
+      expect(logSpy).toHaveBeenCalledWith('Applying proactive rate limit (limiting req/s).');
       expect(rlSpy).toHaveBeenCalled();
     });
   });
diff --git a/packages/rest/src/client.ts b/packages/rest/src/client.ts
index 58ad85e..b27184e 100644
--- a/packages/rest/src/client.ts
+++ b/packages/rest/src/client.ts
@@ -37,7 +37,7 @@ export interface ClientOptions {
    * Whether to slow down API calls so that the Figma REST API's rate limit
    * policy isn't violated. Prevents errors when doing multiple consecutive calls.
    */
-  rateLimit?: boolean;
+  rateLimit?: boolean | 'reactive' | 'proactive';
 }
 
 export type ClientInterface = Omit<ApiInterface<ClientOptions>, 'setSecurityData'> & {
@@ -50,7 +50,7 @@ export async function Client(opts: ClientOptions = {}): Promise<ClientInterface>
     mode = process.env.NODE_ENV,
     oauthToken = process.env.FIGMA_OAUTH_TOKEN,
     personalAccessToken = process.env.FIGMA_PERSONAL_ACCESS_TOKEN,
-    rateLimit = true,
+    rateLimit = 'reactive',
   } = opts;
 
   const isDevelopment = mode === 'development';
@@ -103,13 +103,17 @@ export async function Client(opts: ClientOptions = {}): Promise<ClientInterface>
    * We do this without a loop because of TypeScript limitations,
    * so this must be maintained on every major API version release. */
   if (rateLimit) {
-    log(`Applying rate limit proxy to API client.`);
+    log('Applying rate limit safeguards to API client.');
 
-    api.instance.interceptors.request.use(
-      rateLimitRequestInterceptor(defaultKeyGenerator, cacheInstance?.getCache()),
-    );
+    if (rateLimit === true || rateLimit === 'proactive') {
+      log('Applying proactive rate limit (limiting req/s).');
+      api.instance.interceptors.request.use(
+        rateLimitRequestInterceptor(defaultKeyGenerator, cacheInstance?.getCache()),
+      );
+    }
 
     // Add response interceptor for 429 handling.
+    log('Applying exponential retry on Error 429.');
     const rlConfig = get429Config();
     axiosRetry(api.instance, {
       onMaxRetryTimesExceeded: (error) => {