feat: add flex mode option for cost savings (#372)

Adding in an option to turn on flex processing mode to reduce costs when
running the agent.

Bumped the openai typescript version to add the new feature.

---------

Co-authored-by: Thibault Sottiaux <tibo@openai.com>

codex-cli/src/utils/agent/agent-loop.ts

@@ -516,6 +516,7 @@ export class AgentLoop {
               stream: true,
               parallel_tool_calls: false,
               reasoning,
+              ...(this.config.flexMode ? { service_tier: "flex" } : {}),
               tools: [
                 {
                   type: "function",

codex-cli/src/utils/compact-summary.ts

@@ -9,9 +9,17 @@ import OpenAI from "openai";
  * @param model The model to use for generating the summary
  * @returns A concise structured summary string
  */
+/**
+ * Generate a condensed summary of the conversation items.
+ * @param items The list of conversation items to summarize
+ * @param model The model to use for generating the summary
+ * @param flexMode Whether to use the flex-mode service tier
+ * @returns A concise structured summary string
+ */
 export async function generateCompactSummary(
   items: Array<ResponseItem>,
   model: string,
+  flexMode = false,
 ): Promise<string> {
   const oai = new OpenAI({
     apiKey: process.env["OPENAI_API_KEY"],
@@ -44,6 +52,7 @@ export async function generateCompactSummary(
 
   const response = await oai.chat.completions.create({
     model,
+    ...(flexMode ? { service_tier: "flex" } : {}),
     messages: [
       {
         role: "assistant",

codex-cli/src/utils/config.ts

@@ -79,6 +79,9 @@ export type AppConfig = {
   memory?: MemoryConfig;
   /** Whether to enable desktop notifications for responses */
   notify: boolean;
+
+  /** Enable the "flex-mode" processing mode for supported models (o3, o4-mini) */
+  flexMode?: boolean;
   history?: {
     maxSize: number;
     saveHistory: boolean;