Error 403: what to do when gemini-3.1-pro-preview fails

A 403 means you are authenticated but not allowed: the key is valid, yet it may not call this model or this group.

403 Forbidden means the identity was recognised but access is denied. The difference from 401: 401 asks who you are, 403 says you lack permission. Usually the model needs to be added to the allowed list or group bound to your key.

gemini-3.1-pro-preview is served by Google. Everything on this page — triggers, fixes and measured data — is compiled from the real runtime behaviour of this model at the gateway layer.

At this gateway, the most common trigger is: This key is not authorised for this model. The recommended first action is: Add this model to the key allowlist on the Tokens page.

Common causes

  • This key is not authorised for this model
  • The model is not in the group the key is bound to
  • The key has an IP allowlist that excludes your current IP
  • The model was retired or requires higher permissions

How to fix

  • Add this model to the key allowlist on the Tokens page
  • Confirm the key group includes this model
  • Review the IP allowlist settings
  • Switch to a model you are allowed to call

Retry with exponential backoff

The snippet below retries when gemini-3.1-pro-preview returns 403, up to 5 attempts, with an increasing wait plus random jitter so concurrent calls do not retry in lockstep. Read the base URL and API key from environment variables — never hardcode them.

import os, time, random
import requests

BASE  = os.getenv("OPENAI_BASE_URL")   # e.g. https://<your-gateway>/v1
KEY   = os.getenv("OPENAI_API_KEY")
MODEL = 'gemini-3.1-pro-preview'


def chat(messages, retries=5):
    """Retry with exponential backoff + jitter."""
    for i in range(retries):
        try:
            r = requests.post(
                BASE + "/chat/completions",
                headers={"Authorization": "Bearer " + KEY},
                json={"model": MODEL, "messages": messages, "stream": True},
                timeout=60,
            )
            if r.status_code == 429 or r.status_code >= 500:
                time.sleep(min(2 ** i + random.uniform(0, 1), 30))
                continue
            r.raise_for_status()
            return r.json()
        except requests.exceptions.Timeout:
            time.sleep(min(2 ** i + random.uniform(0, 1), 30))
    raise RuntimeError("gave up after " + str(retries) + " retries")


print(chat([{"role": "user", "content": "hello"}]))

Key facts for this model

TPS3738.12
Avg latency26006 ms
Success rate100%
API endpointhttps://api.airai.cc/v1
OpenAI-compatibleOpenAI-compatible
VendorGoogle
Context1M
CapabilitiesReasoning, Tools, Files, Vision, Audio
API formatsopenai, openai-response, openai-response-compact, anthropic, gemini, openai-alpha-search
Billing formulap * 2 + cr * 0.2 + c * 12) : tier("200k_plus", p * 4 + cr * 0.4 + c * 18

FAQ

How much concurrency is safe?

Lower the concurrency first — most throughput complaints disappear once you do. This model has a 1M context window and comes from Google. Measured throughput is 3738.12, a useful ceiling for concurrency. Add a cache layer so repeated requests do not all hit the model.

Is 403 more likely with direct frontend calls or backend proxying?

Keep a fallback model ready as well. Building a fallback into the architecture is more reliable than patching errors one by one. This model from Google has several upstream nodes the gateway can switch between. Keep a lighter fallback model ready so the main flow never breaks. Put the model name in config, so switching upstreams needs no code change.

Do failed requests count against my rate limit quota?

Only the output already produced is billed; the failed part is not. With billing p * 2 + cr * 0.2 + c * 12) : tier("200k_plus", p * 4 + cr * 0.4 + c * 18, failed requests are not counted toward usage. Check your balance and rate limits in the console before debugging code.

Why does gemini-3.1-pro-preview only return 403 during certain hours?

Group the errors by time and node first; the pattern is usually obvious once you do. If it only happens in production, it is usually an environment difference, not the model. Measured success rate is 100%, and most failures surface as 403. Log the request ID on every failure — it beats the status code when debugging. Reproduce it once in a staging environment with the same request body.

Other errors on this model

Other models with the same error

Back to the model page

Data updated: 2026-10-10 18:35

Technical SupportLive Support
Back to Top