Skip to main content
Google

Gemini 4 Neon

Gemini 4 Neon is a chat model from Google, tracked here with third-party throughput and latency measurements alongside provider pricing.

Input from
$3.00 / 1M tokens
on OpenRouter · 2 providers
Output from
$15.00 / 1M tokens
on OpenRouter

API Pricing

Cheapest input on OpenRouter · cheapest output on OpenRouter
ProviderInput / 1MOutput / 1MCached / 1M
$1.50$7.50$0.075
$3.00$15.00$0.150
$3.00$15.00-

Prices updated daily. Last check: Nov 19, 2026

Compare API pricing for every Gemini model →

Gemini 4 Neon pricing by provider

Input, output, and batch rates, plus alternatives, for one provider at a time.

Performance & Benchmarks

Source: Artificial Analysis →
Intelligence
61.4 / 100
Output Speed
136 t/s
Latency (TTFT)
124.9s

Benchmarks measured Nov 2026. Scores are independent evaluations, not vendor-reported.

Model Details

General

Creator
Google
Modalities
Text

Capabilities

Strengths & Limitations

Strengths

  • Measured output throughput of approximately 151.5 tokens per second in Artificial Analysis testing, which supports responsive streaming once generation begins
  • Developed by Google, a provider with established model-serving infrastructure and API tooling
  • Listed as a chat model, so it fits standard conversational and assistant-style request formats
  • Independent third-party latency and throughput measurements are available, rather than only vendor-reported figures
  • Tracked across the providers in our pricing table, allowing direct rate comparison where multiple hosts offer it

Limitations

  • Time to first token measured at roughly 12,360 ms, a long initial delay that is poorly suited to latency-sensitive interactive use
  • Context window size is not tracked in our database, so maximum input length cannot be confirmed here
  • Supported input and output modalities are not recorded in our metadata
  • No task-specific benchmark scores (coding, math, reasoning) are available in our dataset for this model
  • Published throughput and latency figures come from a single measurement source and may not match what a given provider delivers under load

Key Features

•Chat/conversational completion interface
•Measured output generation rate of ~151.5 tokens per second (Artificial Analysis)
•Measured time to first token of ~12,360 ms (Artificial Analysis)
•Part of Google's Gemini model line
•Streaming-style token output suited to progressive rendering in UIs
•Provider pricing tracked and compared on this page

About Gemini 4 Neon

Gemini 4 Neon is a conversational large language model developed by Google and listed in our catalog as a chat model. Our database records limited verified metadata for it, so this entry focuses on what we can confirm: its creator, its model type, and independent performance measurements rather than unverified capability claims. On measured runtime behavior, Artificial Analysis reports an output rate of roughly 151.5 tokens per second for Gemini 4 Neon, with a time to first token of about 12,360 ms. That combination — a fairly high sustained generation rate paired with a long initial delay — is the pattern typically seen when a model performs substantial internal processing before emitting its first visible token. In practice it means the model streams quickly once it starts, but users should expect a noticeable pause at the beginning of each response. These figures come from a single third-party harness and will vary by provider, region, prompt length, and load. We do not currently track a confirmed context window, modality list, or benchmark suite for Gemini 4 Neon, so this page does not assert those details. For availability and current rates across the providers we monitor, consult the pricing table on this page; serving characteristics such as streaming support, rate limits, and quantization can differ between hosts even for the same model name.

Common Use Cases

Gemini 4 Neon's measured profile — a long time to first token followed by fast sustained generation — points toward workloads where total completion quality matters more than immediate responsiveness. That includes batch or background jobs such as document drafting, summarization pipelines, content generation queues, and asynchronous agent steps where a ten-second-plus startup delay is absorbed by the surrounding workflow. It is a weaker fit for real-time chat widgets, voice interfaces, or autocomplete-style features where users perceive the first-token delay directly. Because our metadata does not confirm the model's context window or modality support, teams planning long-document or multimodal workloads should verify those limits with the specific provider before committing, and should benchmark against their own prompts rather than relying on a single third-party latency figure.

Frequently Asked Questions

How much does Gemini 4 Neon cost?

Pricing varies by provider and by pricing type — input tokens, output tokens, cached input, and batch rates are often billed differently, and the same model can carry different rates on different hosts. Check the pricing table on this page for the current tracked rates.

What is Gemini 4 Neon best used for?

Its measured behavior — about 151.5 output tokens per second after a time to first token near 12.4 seconds — suits asynchronous and batch-oriented text generation: summarization pipelines, long-form drafting, report generation, and background agent steps. Interactive, latency-sensitive interfaces are a weaker match because of the initial delay.

Why is the time to first token so high?

A multi-second time to first token alongside a healthy streaming rate usually indicates the model does meaningful processing before emitting output. We report the measured figure (roughly 12,360 ms from Artificial Analysis) without asserting the underlying cause, since our database does not record architectural details for this model.

What context window does Gemini 4 Neon support?

We do not currently track a confirmed context window for Gemini 4 Neon, so we do not list one here. Confirm the maximum input length with the specific provider you plan to use, as serving limits can differ between hosts of the same model.

Does Gemini 4 Neon accept image or audio input?

Our metadata does not record the supported modalities for this model, so we cannot confirm either way. Check the provider's model documentation before building a multimodal workflow around it.