Flex inference

The Gemini Flex API is an inference tier that offers a 50% cost reduction compared to standard rates, in exchange for variable latency and best-effort availability. It's designed for latency-tolerant workloads that require synchronous processing but don't need the real-time performance of the standard API.

How to use Flex

To use the Flex tier, specify the service_tier as flex in the request body. By default, requests use the standard tier if this field is omitted.

Python

import google.genai as genai

client = genai.Client()

try:
    response = client.models.generate_content(
        model="gemini-3-flash-preview",
        contents="Analyze this dataset for trends...",
        config={'service_tier': 'flex'},
    )
    print(response.text)
except Exception as e:
    print(f"Flex request failed: {e}")

JavaScript

import {GoogleGenAI} from '@google/genai';

const ai = new GoogleGenAI({});

async function main() {
  try {
    const response = await ai.models.generateContent({
      model: "gemini-3-flash-preview",
      contents: "Analyze this dataset for trends...",
      config: { serviceTier: "flex" },
    });
    console.log(response.text);
  } catch (e) {
    console.log(`Flex request failed: ${e}`);
  }
}

await main();

Go

package main

import (
    "context"
    "fmt"
    "log"
    "google.golang.org/genai"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    result, err := client.Models.GenerateContent(
        ctx,
        "gemini-3-flash-preview",
        genai.Text("Analyze this dataset for trends..."),
        &genai.GenerateContentConfig{
            ServiceTier: "flex",
        },
    )
    if err != nil {
        log.Printf("Flex request failed: %v", err)
        return
    }
    fmt.Println(result.Text())
}

REST

"https://generativelanguage.googleapis.com/v1beta/models/gemini-3-flash-preview:generateContent?key=$GOOGLE_API_KEY" \
-H 'Content-Type: application/json' \
-d '{
  "contents": [{
    "parts":[{"text": "Summarize the latest research on quantum computing."}]
  }],
  "service_tier": "FLEX"
}'

How Flex inference works

Gemini Flex inference bridges the gap between the standard API and the 24-hour turnaround of the Batch API. It utilizes off-peak, "sheddable" compute capacity to provide a cost-effective solution for background tasks and sequential workflows.

Feature Flex Priority Standard Batch
Pricing 50% discount 75-100% more than Standard Full price 50% discount
Latency Minutes (1–15 min target) Low (Seconds) Seconds to minutes Up to 24 hours
Reliability Best-effort (Sheddable) High (Non-sheddable) High / Medium-high High (for throughput)
Interface Synchronous Synchronous Synchronous Asynchronous

Key benefits

  • Cost efficiency: Substantial savings for non-production evals, background agents, and data enrichment.
  • Low friction: No need to manage batch objects, job IDs, or polling; simply add a single parameter to your existing requests.
  • Synchronous workflows: Ideal for sequential API chains where the next request depends on the output of the previous one, making it more flexible than Batch for agentic workflows.

Use cases

  • Offline evaluations: Running "LLM-as-a-judge" regression tests or leaderboards.
  • Background agents: Sequential tasks like CRM updates, profile building, or content moderation where minutes of delay are acceptable.
  • Budget-constrained research: Academic experiments that require high token volume on a limited budget.

Rate limits

Flex inference traffic counts towards your general rate limits; it doesn't offer extended rate limits like the Batch API.

Sheddable capacity

Flex traffic is treated with lower priority. If there is a spike in standard traffic, Flex requests may be preempted or evicted to ensure capacity for high-priority users. If you're looking for high-priority inference, check Priority inference

Error codes

When Flex capacity is unavailable or the system is congested, the API will return standard error codes:

  • 503 Service Unavailable: The system is currently at capacity.
  • 429 Too Many Requests: Rate limits or resource exhaustion.

Client responsibility

  • No server-side fallback: To prevent unexpected charges, the system won't automatically upgrade a Flex request to the Standard tier if Flex capacity is full.
  • Retries: You must implement your own client-side retry logic with exponential backoff.
  • Timeouts: Because Flex requests may sit in a queue, we recommend increasing client-side timeouts to 10 minutes or more to avoid premature connection closure.

Adjust timeout windows

You can configure per-request timeouts for the REST API and client libraries, and global timeouts only when using the client libraries.

Always ensure your client-side timeout covers the intended server patience window (e.g., 600s+ for Flex wait queues). The SDKs expect timeout values in milliseconds.

Per-request timeouts

Python

from google import genai

client = genai.Client()

try:
    response = client.models.generate_content(
        model="gemini-3-flash-preview",
        contents="why is the sky blue?",
        config={
            "service_tier": "flex",
            "http_options": {"timeout": 900000}
        },
    )
except Exception as e:
    print(f"Flex request failed: {e}")

# Example with streaming
try:
    response = client.models.generate_content_stream(
        model="gemini-3-flash-preview",
        contents=["List 5 ideas for a sci-fi movie."],
        config={
            "service_tier": "flex",
            "http_options": {"timeout": 60000}
        }
        # Per-request timeout for the streaming operation
    )
    for chunk in response:
        print(chunk.text, end="")

except Exception as e:
    print(f"An error occurred during streaming: {e}")

JavaScript

 import {GoogleGenAI} from '@google/genai';

 const client = new GoogleGenAI({});

 async function main() {
     try {
         const response = await client.models.generateContent({
             model: "gemini-3-flash-preview",
             contents: "why is the sky blue?",
             config: {
               serviceTier: "flex",
               httpOptions: {timeout: 900000}
             },
         });
     } catch (e) {
         console.log(`Flex request failed: ${e}`);
     }

     // Example with streaming
     try {
         const response = await client.models.generateContentStream({
             model: "gemini-3-flash-preview",
             contents: ["List 5 ideas for a sci-fi movie."],
             config: {
                 serviceTier: "flex",
                 httpOptions: {timeout: 60000}
             },
         });
         for await (const chunk of response.stream) {
             process.stdout.write(chunk.text());
         }
     } catch (e) {
         console.log(`An error occurred during streaming: ${e}`);
     }
 }

 await main();

Go

package main

import (
    "context"
    "fmt"
    "log"
    "time"

    "google.golang.org/api/iterator"
    "google.golang.org/genai"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }
    defer client.Close()

    timeoutCtx, cancel := context.WithTimeout(ctx, 900*time.Second)
    defer cancel()

    _, err = client.Models.GenerateContent(
        timeoutCtx,
        "gemini-3-flash-preview",
        genai.Text("why is the sky blue?"),
        &genai.GenerateContentConfig{
            ServiceTier: "flex",
        },
    )
    if err != nil {
        fmt.Printf("Flex request failed: %v\n", err)
    }

    // Example with streaming
    streamTimeoutCtx, streamCancel := context.WithTimeout(ctx, 60*time.Second)
    defer streamCancel()

    iter := client.Models.GenerateContentStream(
        streamTimeoutCtx,
        "gemini-3-flash-preview",
        genai.Text("List 5 ideas for a sci-fi movie."),
        &genai.GenerateContentConfig{
            ServiceTier: "flex",
        },
    )
    for {
        response, err := iter.Next()
        if err == iterator.Done {
            break
        }
        if err != nil {
            fmt.Printf("An error occurred during streaming: %v\n", err)
            break
        }
        fmt.Print(response.Candidates[0].Content.Parts[0])
    }
}

REST

When making REST calls, you can control timeouts using a combination of HTTP headers and curl options:

  • X-Server-Timeout header (server-side timeout): This header suggests a preferred timeout duration (default 600s) to the Gemini API server. The server will attempt to respect this, but it's not guaranteed. The value should be in seconds.

  • --max-time in curl (Client-Side Timeout): The curl --max-time <seconds> option sets a hard limit on the total time (in seconds) that curl will wait for the entire operation to complete. This is a client-side safeguard.

 # Set a server timeout hint of 120 seconds and a client-side curl timeout of 125 seconds.
 curl --max-time 125 \
   -X POST "https://generativelanguage.googleapis.com/v1beta/models/gemini-3-flash-preview:generateContent" \
   -H "Content-Type: application/json" \
   -H "x-goog-api-key: YOUR_API_KEY" \
   -H "X-Server-Timeout: 120" \
   -d '{
   "contents": [{
     "parts":[{"text": "Summarize the latest research on quantum computing."}]
   }],
   "service_tier": "SERVICE_TIER_FLEX"
 }'

Global timeouts

If you want all API calls made through a specific genai.Client instance (client libraries only) to have a default timeout, you can configure this when initializing the client using http_options and genai.types.HttpOptions.

Python

from google import genai

global_timeout_ms = 120000

client_with_global_timeout = genai.Client(
    http_options=types.HttpOptions(timeout=global_timeout_ms)
)

try:
    # Calling generate_content using global timeout...
    response = client_with_global_timeout.models.generate_content(
        model="gemini-3-flash-preview",
        contents="Summarize the history of AI development since 2000.",
        config={"service_tier": "flex"},
    )
    print(response.text)

    # A per-request timeout will *override* the global timeout for that specific call.
    shorter_timeout = 30000
    response = client_with_global_timeout.models.generate_content(
        model="gemini-3-flash-preview",
        contents="Provide a very brief definition of machine learning.",
        config={
            "service_tier": "flex",
            "http_options":{"timeout": shorter_timeout}
        }  # Overrides the global timeout
    )

    print(response.text)

except TimeoutError:
    print(
        f"A GenerateContent call timed out. Check if the global or per-request timeout was exceeded."
    )
except Exception as e:
    print(f"An error occurred: {e}")

JavaScript

import {GoogleGenAI} from '@google/genai';

const globalTimeoutMs = 120000;

const clientWithGlobalTimeout = new GoogleGenAI({httpOptions: {timeout: globalTimeoutMs}});

async function main() {
    try {
        // Calling generate_content using global timeout...
        const response1 = await clientWithGlobalTimeout.models.generateContent({
            model: "gemini-3-flash-preview",
            contents: "Summarize the history of AI development since 2000.",
            config: { serviceTier: "flex" },
        });
        console.log(response1.text());

        // A per-request timeout will *override* the global timeout for that specific call.
        const shorterTimeout = 30000;
        const response2 = await clientWithGlobalTimeout.models.generateContent({
            model: "gemini-3-flash-preview",
            contents: "Provide a very brief definition of machine learning.",
            config: {
                serviceTier: "flex",
                httpOptions: {timeout: shorterTimeout}
            }  // Overrides the global timeout
        });

        console.log(response2.text());

    } catch (e) {
        if (e.name === 'TimeoutError' || e.message?.includes('timeout')) {
            console.log(
                "A GenerateContent call timed out. Check if the global or per-request timeout was exceeded."
            );
        } else {
            console.log(`An error occurred: ${e}`);
        }
    }
}

await main();

Go

 package main

 import (
     "context"
     "fmt"
     "log"
     "time"

     "google.golang.org/genai"
 )

 func main() {
     ctx := context.Background()
     client, err := genai.NewClient(ctx, nil)
     if err != nil {
         log.Fatal(err)
     }
     defer client.Close()

     model := client.GenerativeModel("gemini-3-flash-preview")

     // Go uses context for timeouts, not client options.
     // Set a default timeout for requests.
     globalTimeout := 120 * time.Second
     fmt.Printf("Using default timeout of %v seconds.\n", globalTimeout.Seconds())

     fmt.Println("Calling generate_content (using default timeout)...")
     ctx1, cancel1 := context.WithTimeout(ctx, globalTimeout)
     defer cancel1()
     resp1, err := model.GenerateContent(ctx1, genai.Text("Summarize the history of AI development since 2000."), &genai.GenerateContentConfig{ServiceTier: "flex"})
     if err != nil {
         log.Printf("Request 1 failed: %v", err)
     } else {
         fmt.Println("GenerateContent 1 successful.")
         fmt.Println(resp1.Text())
     }

     // A different timeout can be used for other requests.
     shorterTimeout := 30 * time.Second
     fmt.Printf("\nCalling generate_content with a shorter timeout of %v seconds...\n", shorterTimeout.Seconds())
     ctx2, cancel2 := context.WithTimeout(ctx, shorterTimeout)
     defer cancel2()
     resp2, err := model.GenerateContent(ctx2, genai.Text("Provide a very brief definition of machine learning."), &genai.GenerateContentConfig{
         ServiceTier: "flex",
     })
     if err != nil {
         log.Printf("Request 2 failed: %v", err)
     } else {
         fmt.Println("GenerateContent 2 successful.")
         fmt.Println(resp2.Text())
     }
 }

Implement retries

Because Flex is sheddable and fails with 503 errors, here is an example of optionally implementing retry logic to continue with failed requests:

Python

import time
from google import genai

client = genai.Client()

def call_with_retry(max_retries=3, base_delay=5):
    for attempt in range(max_retries):
        try:
            return client.models.generate_content(
                model="gemini-3-flash-preview",
                contents="Analyze this batch statement.",
                config={"service_tier": "flex"},
            )
        except Exception as e:
            # Check for 503 Service Unavailable or 429 Rate Limits
      print(e.code)
            if attempt < max_retries - 1:
                delay = base_delay * (2 ** attempt) # Exponential Backoff
                print(f"Flex busy, retrying in {delay}s...")
                time.sleep(delay)
            else:
                # Fallback to standard on last strike (Optional)
                print("Flex exhausted, falling back to Standard...")
                return client.models.generate_content(
                    model="gemini-3-flash-preview",
                    contents="Analyze this batch statement."
                )

# Usage
response = call_with_retry()
print(response.text)

JavaScript

 import {GoogleGenAI} from '@google/genai';

 const ai = new GoogleGenAI({});

 async function sleep(ms) {
   return new Promise(resolve => setTimeout(resolve, ms));
 }

 async function callWithRetry(maxRetries = 3, baseDelay = 5) {
   for (let attempt = 0; attempt < maxRetries; attempt++) {
     try {
       console.log(`Attempt ${attempt + 1}: Calling Flex tier...`);
       const response = await ai.models.generateContent({
         model: "gemini-3-flash-preview",
         contents: "Analyze this batch statement.",
         config: { serviceTier: 'flex' },
       });
       return response;
     } catch (e) {
       if (attempt < maxRetries - 1) {
         const delay = baseDelay * (2 ** attempt);
         console.log(`Flex busy, retrying in ${delay}s...`);
         await sleep(delay * 1000);
       } else {
         console.log("Flex exhausted, falling back to Standard...");
         return await ai.models.generateContent({
           model: "gemini-3-flash-preview",
           contents: "Analyze this batch statement.",
         });
       }
     }
   }
 }

 async function main() {
     const response = await callWithRetry();
     console.log(response.text);
 }

 await main();

Go

 package main

 import (
     "context"
     "fmt"
     "log"
     "math"
     "time"

     "google.golang.org/genai"
 )

 func callWithRetry(ctx context.Context, client *genai.Client, maxRetries int, baseDelay time.Duration) (*genai.GenerateContentResponse, error) {
     modelName := "gemini-3-flash-preview"
     content := genai.Text("Analyze this batch statement.")
     flexConfig := &genai.GenerateContentConfig{
         ServiceTier: "flex",
     }

     for attempt := 0; attempt < maxRetries; attempt++ {
         log.Printf("Attempt %d: Calling Flex tier...", attempt+1)
         resp, err := client.Models.GenerateContent(ctx, modelName, content, flexConfig)
         if err == nil {
             return resp, nil
         }

         log.Printf("Attempt %d failed: %v", attempt+1, err)

         if attempt < maxRetries-1 {
             delay := time.Duration(float64(baseDelay) * math.Pow(2, float64(attempt)))
             log.Printf("Flex busy, retrying in %v...", delay)
             time.Sleep(delay)
         } else {
             log.Println("Flex exhausted, falling back to Standard...")
             return client.Models.GenerateContent(ctx, modelName, content)
         }
     }
     return nil, fmt.Errorf("retries exhausted") // Should not be reached
 }

 func main() {
     ctx := context.Background()
     client, err := genai.NewClient(ctx, nil)
     if err != nil {
         log.Fatal(err)
     }
     defer client.Close()

     resp, err := callWithRetry(ctx, client, 3, 5*time.Second)
     if err != nil {
         log.Fatalf("Failed after retries: %v", err)
     }
     fmt.Println(resp.Text())
 }

Pricing

Flex inference is priced at 50% of the standard API and billed per token.

Supported models

The following models support Flex inference:

Model Flex inference
Gemini 3.1 Flash-Lite Preview ✔️
Gemini 3.1 Pro Preview ✔️
Gemini 3 Flash Preview ✔️
Gemini 3 Pro Image Preview ✔️
Gemini 2.5 Pro ✔️
Gemini 2.5 Flash ✔️
Gemini 2.5 Flash Image ✔️
Gemini 2.5 Flash-Lite ✔️

What's next

Read about Gemini's other inference and optimization options: