Gemini 4 Neon API pricing per million tokens

Gemini 4 Neon costs $3 per million input tokens and $15 per million output, with a 1.05M context window. Through AIGROW API credit, it costs $1.50 and $7.50.

$25 a month brings $110 of API credit and cancels any time; $10 once buys $40. US dollars per million tokens, standard rate, shortest context band. Number 3 of 426 on LMArena’s blind votes, published 19 November 2026.

$3 / $15Gemini 4 Neon · list priceper million tokens, input / output
1,000 support replies
$15.00
Cached input
$0.15
Context window
1,048,576 tokens
Released
18 Nov 2026
Weights
Closed weights
From Google’s own pages, 18 November 2026.

Gemini 4 Neon costs $3 per million input tokens and $15 per million output tokens on Google’s own API, and $0.15 for cached input. At those prices 1,000 support replies cost $15.00, the 6th most expensive of 36 priced models on our list.

Gemini 4 Neon through AIGROW API credit

AIGROW API credit runs Gemini 4 Neon at $1.50 / $7.50 per million input / output tokens, 50% below list price.

Cache and long prompts
Cache reads cost $0.075 per million tokens through AIGROW credit and $0.15 at Google’s list price. Through AIGROW credit, a request to Gemini 4 Neon bills at the same rates whatever the length of its prompt, up to the 1,048,576-token context window.
Verified genuine by AIGROW
Gemini 4 Neon passed AIGROW’s own check against the same model at its maker on 21 November 2026: both got the same prompts, and their scores on questions with one right answer, their prompt token counts and their tool calls were compared. Google has not reviewed or endorsed AIGROW or this check.
How to pay
Pay at the Stripe checkout by card, Apple Pay, Google Pay or Amazon Pay. A one-off pack can also be paid with WeChat Pay, with Pix in Brazil, with UPI in India, or in crypto on the price list. The redemption code is emailed the moment payment clears. Sign in, paste the code into API access and create a key that starts with ag_. The code is not tied to the email you paid with. No Google account is needed.

What a job costs

Three jobs, 1,000 runs of each, at the list price above, and where each puts Gemini 4 Neon among the 36 current models with a list price.

Gemini 4 Neon: cost of 1,000 runs of three jobs
Job1,000 runsAmong 36
Support reply3,000 in, 400 out.$15.006th most expensive
Document summary25,000 in, 1,000 out.$90.006th most expensive
Agent task80,000 in (70,000 cached), 3,000 out.$85.506th most expensive

The token counts are ours, chosen to look like real work; the prices are Google’s. All three jobs stay under 100,000 tokens a request, below every long-context surcharge on the list.

The small print

What the base rate above leaves out, copied from the maker’s pages: surcharges, discounts, promotions and cache pricing.

Google places it between Gemini 4 Hydrogen and Gemini 4 Carbon, for complex software engineering, knowledge work and science, and its model card says it is adapted from Gemini 4 Carbon. Output includes thinking tokens. Implicit caching only. One rate across the full 1M context, with no long-context premium. Batch and Flex 50% off; Priority $5.40 / $27. Not on the Free tier. Max output 65,536, or up to 1,048,576 with long decode.

Gemini 4 Neon specifications

The limits and features Google documents for Gemini 4 Neon, read on 18 November 2026.

Maximum output
65,536 tokens
Knowledge cutoff
Mar 2026
Input types
Text, images, PDF files, video and audio
Reasoning
Always on. Effort levels: low, medium, high, xhigh
Function calling
Yes
Structured output
Yes
Web search
Yes
Code execution
Yes
Computer use
Yes
Prompt caching
Automatic
Batch discount
50% off
Status
Current
Fine-tuning
Not offered on Google’s API

Every fact above was read on these pages, and a fact they do not state is left out:aistudio.google.com/docs/models/gemini-4-neon.mdaistudio.google.com/docs/whats-new-gemini-4-neon.mddeepmind.google/models/model-cards/gemini-4-neon/docs.cloud.google.com/gemini-enterprise-agent-platform/models/gemini/4-neonaistudio.google.com/docs/pricing.md

Gemini 4 Neon against the models it is weighed with

Each comparison puts both list prices side by side and costs the same three jobs on each.

vs Gemini 4 Carbon
Gemini 4 Neon against the Gemini 4 Carbon it is adapted from, at 1.5 times the Carbon price through November 2026 and 75% of it from 1 December.
vs Gemini 4 Hydrogen
Gemini 4 Neon against Gemini 4 Hydrogen, Google's default for everyday coding, at twice the Hydrogen price on input, output and cache reads.
vs Gemini 4 Argon
Google's Gemini 4 Neon against Gemini 4 Argon, at 1.5 times the Argon price through November 2026 and 75% of it from 1 December, when Argon doubles.
vs Gemini 3.8 Flash
Google's Gemini 4 Neon against Gemini 3.8 Flash, at four times the Flash input and output price through 2026 and twice it from 2027.
vs Claude Sonnet 5.5
Google's Gemini 4 Neon against Anthropic's Sonnet 5.5, at 1.5 times the Sonnet price on input, output and cache reads.
vs Claude Opus 5.5
Google's Gemini 4 Neon against Anthropic's Opus 5.5, at three quarters of the Opus price on input, output and cache reads.
vs Claude Fable 5.2
Google's Gemini 4 Neon against Anthropic's Fable 5.2, at 37.5% of the Fable price on input and output and three quarters of its cache read.
vs GPT-6.2 Sol
Google's Gemini 4 Neon against OpenAI's GPT-6.2 Sol, at 1.5 times the Sol price, with one Neon rate across 1M tokens where Sol's rises above 272K.
vs GPT-6.2 Astra
Google's Gemini 4 Neon against OpenAI's GPT-6.2 Astra, at 37.5% of the Astra price on input, output and cache reads.
vs Grok 4.7
Google's Gemini 4 Neon against xAI's Grok 4.7, at 1.5 times the Grok input price and 2.5 times its output, with a cache read at 30% of Grok's.
vs DeepSeek V4 Pro
Google's Gemini 4 Neon against DeepSeek's open-weight V4 Pro, at over twice the DeepSeek peak input price and nearly four times its peak output.
vs Qwen4-Max
Google's Gemini 4 Neon against Alibaba's Qwen4-Max, at twice the Qwen input and cache-read price and over three times its output price.
vs Kimi K3.1 Pro
Google's Gemini 4 Neon against Moonshot's open-weight Kimi K3.1 Pro, at 1.5 times the Kimi price, with a cache read three quarters of Kimi's.

Other Google models

Other Google models
ModelInputOutputContext1,000 replies
Gemini 3.5 Flash-LiteGoogle$0.30$2.501.05M$1.90
Gemini 3.8 FlashGoogle$0.75$3.751.05M$3.75
Gemini 4 HydrogenGoogle$1.50$7.501.05M$7.50
Gemini 4 CarbonGoogle$2.00$10.001.05M$10.00
Gemini 4 ArgonGoogle$2.00$10.001.05M$10.00
Gemini 3.1 Flash-LiteGoogleReplaced by Gemini 3.5 Flash-Lite$0.25$1.501.05M$1.35
Gemini 3 FlashGoogleReplaced by Gemini 3.8 Flash$0.50$3.001.05M$2.70
Gemini 3.1 Pro (preview)GoogleReplaced by Gemini 4 Argon$2.00$12.001.05M$10.80

Where the price comes from

Read on 18 November 2026. Each price is the standard rate for the shortest context band, with no batch or priority discount; long-context surcharges, batch rates and promotions are in each model’s small print. No price was taken from a reseller.

Run Gemini 4 Neon up to 50% below list price

Through AIGROW API credit, Gemini 4 Neon costs $1.50 / $7.50 per million input / output tokens against $3.00 / $15.00 at list, and $10 buys $40 of credit. The model id is gemini-4-neon.

Prices checked 18 November 2026 · See them by release date