Gemini 3.8 Flash API: Pricing, Intro Rate and Setup

Gemini 3.8 Flash is Google's current Flash model, announced on September 2, 2026. The API model ID is gemini-3.8-flash, the context window is 1,000,000 tokens, and the introductory list price is $0.75 per 1M input tokens and $3.75 per 1M output tokens through 31 December 2026. On 1 January 2027 that becomes $1.50 / $7.50.

It is live on RouterPlex now, through the same OpenAI-compatible endpoint used for GPT, Claude, Grok and the rest of the catalog.

Gemini 3.8 Flash API pricing #

A $5 prepaid balance at the intro rate buys roughly 6.7 million input tokens, or 1.3 million output tokens, if a workload used only one category. Real requests mix both.

Google also publishes discounted cached and batch halves on the Gemini API. RouterPlex bills the standard intro input and output rates in the table. Do not assume a cache or batch discount on this key unless the live catalog says otherwise.

3.6, 3.7 and 3.8 Flash share the intro rate #

The headline that people miss: Gemini 3.6 Flash, Gemini 3.7 Flash and Gemini 3.8 Flash are the same price right now. All three sit on Google's introductory $0.75 / $3.75 window through the end of 2026.

gemini-3-flash is a different model ID from 3.8 Flash, at a lower rate. If a comparison table or a chat UI just says "Gemini 3 Flash", check the string. On RouterPlex the ids are explicit.

Google announced Gemini 3.8 Flash Cyber on the same day. RouterPlex does not serve Flash Cyber. The route here is gemini-3.8-flash only.

Live pages: Gemini 3.8 Flash, Gemini 3.7 Flash, Gemini 3.6 Flash, Gemini 3 Flash.

Cost per agent turn #

One 40,000-token prompt returning 2,000 tokens, no cache:

The arithmetic for the 3.8 Flash intro row: 40,000 × $0.75 / 1M + 2,000 × $3.75 / 1M = $0.030 + $0.0075 = $0.0375. After 1 January 2027 that doubles on both sides.

If you are buying Flash for volume, 2.5 Flash Lite is still the cheapest Gemini on this catalog. 3.8 Flash is the current head of the line at a still-low intro rate, not the lowest Gemini token price.

Gemini 3.8 Flash specifications #

Call Gemini 3.8 Flash with RouterPlex #

curl https://api.routerplex.com/v1/chat/completions \
  -H "Authorization: Bearer $ROUTERPLEX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.8-flash",
    "messages": [
      {"role": "user", "content": "Summarise the risk register in this incident report."}
    ]
  }'

Python uses the same OpenAI client as every other chat model on this key:

import os
from openai import OpenAI
 
client = OpenAI(
    api_key=os.environ["ROUTERPLEX_API_KEY"],
    base_url="https://api.routerplex.com/v1",
)
 
response = client.chat.completions.create(
    model="gemini-3.8-flash",
    messages=[{"role": "user", "content": "Extract the action items from these notes."}],
)
 
print(response.choices[0].message.content)

Claude Code can point at the same model with ANTHROPIC_BASE_URL=https://api.routerplex.com and model gemini-3.8-flash. See the Claude Code setup guide.

Which Flash should you actually call? #

• Current Flash quality at intro price — gemini-3.8-flash. Same dollars as 3.6 and 3.7 until 31 December 2026, newer weights.
• Cheaper than the 3.6–3.8 intro rate — gemini-3-flash at $0.50 / $3.00, or gemini-2.5-flash at $0.30 / $2.50.
• Volume / classification — gemini-2.5-flash-lite at $0.10 / $0.40.
• Need a flagship, not Flash — GPT-6 Astra or Grok 4.6, on the same key.

Give the Flash key its own hard budget. Intro rates expire on a calendar date, not when you notice the invoice. The live Gemini 3.8 Flash page is the number this route will actually bill.