Gemini 4 Neon API Pricing
Gemini 4 Neon is the middle model of Google's Gemini 4 line, generally available on the Gemini API since November 18, 2026. It costs $3.00/M input, $15.00/M output and $0.15/M cached input, one rate across the whole 1M-token window, with Batch and Flex at half price and no free tier. Google places it between Gemini 4 Hydrogen and Gemini 4 Carbon for complex coding, knowledge work and science that runs often. Checked by hand against Google's Gemini API pricing page.
Gemini 4 Neon API cost calculator.
Live calculator on Google's paid-tier rate, which has no end date and no long-context step. Neon defaults to the high thinking level; dropping to medium or low cuts the output share, and xhigh raises it.
Gemini 4 Neon cost on real workloads.
Multi-file refactor step
Planner turn over parallel subagents
Spreadsheet and filing analysis
Find-and-patch vulnerability run
This is a chars-per-token approximation, not a real tokenizer. Actual tokens vary by language, code density, and tool-call overhead — counts are typically ±10–20% off for English prose, more for code or non-Latin scripts. For exact billing, use the vendor's official tokenizer.
| Model | Input /M | Output /M | Effective blended | Context | Best for |
|---|---|---|---|---|---|
| Gemini 4 Neon Current | $3.00 cache $0.15 | $15.00 | $1.81 agentic 92/8 | 1M | Mid-tier Gemini 4 for complex work that runs often |
| Gemini 4 Carbon | $2.00 cache $0.10 | $10.00 | $1.21 −33% · promo | 1M | Google's most capable model |
| Gemini 4 Hydrogen | $1.50 cache $0.075 | $7.50 | $0.905 −2.0× | 1M | Fast, everyday Gemini 4 with a free tier |
| Gemini 4 Argon | $2.00 cache $0.10 | $10.00 | $1.21 −33% · promo | 1M | Gemini 4 frontier, fewer thinking tokens than Carbon |
| Gemini 3.5 Flash-Lite | $0.30 cache $0.03 | $2.50 | $0.272 −6.6× | 1M | High-volume, simpler tasks |
| Claude Opus 5.5 | $4.00 cache $0.20 | $20.00 | $2.58 +42% | 1M | Anthropic's frontier workhorse |
| GPT-6.2 Sol | $2.00 cache $0.10 | $10.00 | $1.29 −29% | 1.05M | OpenAI's GPT-6.2 workhorse |
| Claude Sonnet 5.5 | $2.00 cache $0.10 | $10.00 | $1.29 −29% | 1M | Anthropic's mid tier for agent loops |
Compare siblings
Gemini 4 Neon has published scores on 6 benchmarks we track, and leads 2 of them. All of them are the lab's own reports; each row links to the page that published it.
| Benchmark | Score | Rank | Leader | Source |
|---|---|---|---|---|
| SEC-bench Pro Cybersecurity | 90.5% as reported | #1 of 11 | Gemini 4 Neon leads | Vendor-reported Google DeepMind |
| AutomationBench Agents and computer use · Main set (as reported) | 56.3% as reported | #3 of 23 | MiniMax M3.1 Flash · 62.8% | Vendor-reported Google DeepMind |
| Terminal-Bench-Science AI research | 70.5% as reported | #3 of 14 | Claude Mythos 5.2 · 75.3% | Vendor-reported Google DeepMind |
| Chartography Vision and documents · Tool use not stated | 73.5% tool use not stated | #1 of 4 | Gemini 4 Neon leads | Vendor-reported Google DeepMind |
| Agents' Last Exam Agents and computer use · ALE v1.1 score | 75.8% v1.1, score | #2 of 3 | Gemini 4 Carbon · 77.4% | Vendor-reported Google DeepMind |
| FrontierCode Coding · FrontierCode 2.0 | 46.2% main score, default high thinking level | #2 of 2 | Claude Fable 5.2 · 49.1% | Vendor-reported Google DeepMind |
Gemini 4 Neon API pricing FAQ.
What Gemini 4 Neon costs, when it is cheaper than Carbon and when it is not, and how it sits next to Hydrogen, Claude Opus 5.5 and GPT-6.2 Sol.
Q · 01 How much does Gemini 4 Neon cost? +
$3.00/M input, $15.00/M output (thinking tokens included) and $0.15/M cached input on the paid tier, flat across the whole 1M-token window with no end date. Under this site's 92/8 agentic blend at an 82% cache-hit rate that is $1.81/M effective.Q · 02 Is Neon cheaper than Gemini 4 Carbon? +
$2.00/M input and $10.00/M output, below Neon's rate. On December 1 Google doubles Carbon's card, and from then Neon is the cheaper of the two by a quarter on every token. Google also says Neon spends fewer tokens per task: at the default high level it puts Neon's cost per task 32% below Carbon's.Q · 03 Neon, Hydrogen or Carbon? +
$1.50/M input, half of Neon's; Carbon is for the hardest architecture, debugging, science and security problems; Neon sits between them for complex work that runs often. Google pairs them too: Neon plans and directs, parallel Hydrogen subagents do the work.Q · 04 Is there a free tier? +
Q · 05 What is the context window? +
1,048,576 tokens shared by input, thinking and output. A synchronous call returns up to 65,536 output tokens; with long decode and background execution one response can run up to 1M tokens minus the input. All of it bills at the output rate, and there is no higher rate for long prompts.Q · 06 Are there Batch, Flex or Priority rates? +
$1.50/M input and $7.50/M output. Priority is 1.8x: $5.40/M input and $27.00/M output. Caching is implicit only, with no storage charge, for prompts of at least 4,096 tokens; explicit caching is not supported.Q · 07 How does Neon compare on price with Claude Opus 5.5 and GPT-6.2 Sol? +
$1.81/M, against $2.58/M for Claude Opus 5.5 and $1.29/M for GPT-6.2 Sol, the two models Google compares it with on its model card.