Gemini 4 Neon is here. When is it worth switching models?
Compare Gemini 4 Neon in two coding setups, work through a $0.45 API example, and find Neon in Gemini, Google AI Studio, and Antigravity.
On this page

Some bugs look like ten-minute fixes until you've traced them through three files and still haven't found the cause. Say an order export occasionally puts a duplicate row in the CSV, but only after the job times out and retries. Filtering out duplicates might tidy up the export. Can you tell whether the retry also drops an order?
That's the sort of job I'd give Gemini 4 Neon. Follow the query, the job queue, and the CSV writer. Then come back with a test that fails before the fix and passes after it. Google announced Neon on November 18, 2026, between Hydrogen for everyday tasks and Carbon for its hardest work. I'd start in that middle tier for the stubborn jobs that come up every week. Google's launch announcement
Which Gemini 4 Neon coding benchmark would I look at first?
SWE-sweep tells an interesting story. In Google's published results, Neon scored 12.8% with the lightweight Mini-SWE harness and 39.0% with Google Antigravity. Same model, different tools and execution environment. If you use a different editor or a script of your own, try a task in that setup before you plan around either number. Google DeepMind's model card
Take a ticket you know how to verify. Give Neon and your usual model the same starting code and the same instructions. Here's what I'd write for that CSV bug.
The order CSV sometimes contains duplicate rows after an export job times out and retries.
Trace the query, retry, and CSV writer. Show where the duplicate comes from.
Add a test that reproduces the bug. Verify every original order appears exactly once after a retry.
List the files you changed and the tests you ran, with their results.Read the patch and the test afterward. If the fix just filters rows at the end, check what happens to the job's write state on retry. The coding model test template gives you room to record test results, how long it took to get a usable patch, and how often you had to step in.
What does a Gemini 4 Neon API request cost?
The Gemini API's standard rates per million tokens are $3 for input, $0.15 for cached input, and $15 for output. Google bills thinking tokens as output. Here's a worked example at the standard rates. Google AI Studio pricing
| Usage in one request | Cost |
|---|---|
| 100,000 uncached input tokens | $0.30 |
| 10,000 output tokens, including thinking | $0.15 |
| Total | $0.45 |
If that second line grows to 20,000 tokens, the bill reaches $0.60. The answer on screen can still be a few paragraphs long because some of those tokens went into thinking. The API defaults to high and also offers low, medium, and xhigh. Before running a batch of similar tickets, I'd try the default level and see which ones need more effort. Gemini 4 Neon API guide
Where can you use Neon?
- Gemini on the web and in the app. Google AI Pro and Ultra users can pick Gemini 4 Neon from the model menu. Google's help page says
Autochooses between Hydrogen and a lightweight model, so open the menu if you want to try Neon. Gemini Apps Help - Google Antigravity. With an eligible account, use the model menu below the prompt box. Neon is the default for new conversations when you haven't already selected a model. The highest thinking level in the menu is called
Extra. Antigravity model docs - Google AI Studio and the Gemini API. Pro and Ultra users can try Neon in AI Studio. For an API call from your own script, use
gemini-4-neonin a paid-tier Gemini API project. Neon API specifications
Common questions
Can you select Gemini 4 Neon on a free account?
Google currently lists Neon under Pro and Ultra in Gemini, and its Gemini API price sheet lists it on the paid tier. Free Gemini users can still use Auto. Gemini Apps Help
Where do you get the million-token context window?
The Gemini API has a shared window of about a million tokens for input, thinking, and output. Antigravity lists Neon at up to 512K tokens. Check the limit for the product you're using before you hand it a stack of files. Gemini API guide · Antigravity model docs
Can Neon read PDFs and videos?
The Gemini API accepts text, image, audio, video, and PDF input and returns text. If you have two versions of a manual, ask Gemini 4 Neon to list the instructions that changed and give a page number for each. Spot-check a few against the original files. Neon API specifications
Model and workflow guides
GPT-6.2 Astra & Sol Guide: Specs, Pricing & Choice
Beginner guide to OpenAI's GPT-6.2 Astra and Sol: Claude vs DeepSeek price table, plain-English benchmarks, 272K cliff, and step-by-step DSH & Python setup.
Read →Models and costsGPT-6.1 Terra Guide: Uses, Pricing and Sol Comparison
Beginner guide to OpenAI's GPT-6.1 Terra: half-price visual, computer-use, and writing model compared with Sol and DeepSeek, plus DSH and Python setup steps.
Read →Model setup and costsConnect Local Qwen 3.8 27B to DSH
Beginner guide to connecting local Qwen 3.8 27B to DSH: GPU memory table, LM Studio / Ollama / llama.cpp port setup, exact 5-field config, and error fixes.
Read →Model setup and costsCodex Costs: API Billing vs Subscription Usage
Beginner guide to OpenAI Codex costs: ChatGPT Plus login vs API-key billing table, real per-prompt dollar math, and 3 steps to lock a monthly budget cap.
Read →Model setup and costsConnect Grok API to DSH: Setup & Tools
Step-by-step beginner guide to connecting xAI's Grok API (grok-4.7) to DSH: Grok vs Groq comparison, exact 5-field custom provider setup, and 401/404 fixes.
Read →Model setup and costsChoose a Gemini Model by Task and Cost
Beginner guide to Google Gemini model selection: compare 3.8 Flash, 4 Argon, and DeepSeek prices, watch 2027 rate changes, and set up DSH in 4 simple steps.
Read →Model benchmarks and engineeringGemini 4 Argon Coding Benchmarks & 1M Window
Beginner guide to Google's Gemini 4 Argon coding scores vs Claude and GPT, how its shared 1M window and 65K output cap work, plus 4-step DSH setup.
Read →API costs and enterprise ROIGemini 4 Argon API Pricing, Cache & ROI Guide
Beginner guide to Gemini 4 Argon API pricing: Dec 2026 2x rate jump, how 95% auto-caching and thinking tokens bill, DeepSeek cost comparison, and calculator.
Read →Model setup and costsClaude Sonnet vs Opus: Which Should You Use?
Beginner guide to Claude Sonnet 5.5 vs Opus 5.5: side-by-side API prices, per-prompt cost math, clear rules on when to upgrade, and 4-step DSH setup.
Read →Models and costsOpus 5.5 vs Fable 5.2: Which Model Is Right for You?
Beginner guide to Claude Opus 5.5 vs Fable 5.2: side-by-side API prices, coding benchmarks, per-task cost math, 30-day retention rules, and 4-step DSH setup.
Read →Content creationHow to draft a video script with Media Desk for DeepSeek Harness
Turn a phone note into a video script, cover, and storyboard with Media Desk for DeepSeek Harness. See how to install it, set up images, and export a draft.
Read →Mobile and remote accessUse DeepSeek Harness on Your Phone: Beginner Guide
Zero-jargon beginner guide to using DeepSeek Harness on your phone: find your computer IP, connect over home Wi-Fi in 2 minutes, or use Tailscale on 4G/5G.
Read →Model interfaces and workflowsMiniMax H3 vs M-Series in DSH: Which Works?
Beginner guide to MiniMax H3 vs M3 in DSH: why H3 is for video while M3 handles coding, how to avoid region 401 errors, and 4-step DSH setup instructions.
Read →API credentials and troubleshootingKimi API Key Errors: Region, Model and Permissions
Beginner fix guide for Kimi / Moonshot API key errors: match .ai vs .cn URLs, avoid retired model names and K3 Max typos, decode 401/404/429, and set up DSH.
Read →Image editing and toolsAI Image Editing: Prompts, Masks and Review
Beginner guide to AI image editing without ruining your subject: copy our 4-line prompt template, avoid PNG mask traps, inspect edges, and use DSH safely.
Read →