Gemini 4 Neon now live across Google platforms with expanded pricing
11/19/2026 — 11/22, 02:16·22 sources (all)·39 reports (all)·In the 2026-11-22 briefing·In the 2026-11-21 briefing
Key takeaways
- Google DeepMind's mid-tier Gemini 4 Neon is now available across the Gemini API, AI Studio, the Gemini app, and GitHub Copilot, positioned b…
Background & analysis
On November 18, 2026, Google DeepMind released Gemini 4 Neon, a mid-tier model announced in a blog post by Koray Kavukcuoglu, SVP of Google DeepMind and Google's chief AI architect. It is the fourth element-named model in the Gemini 4 family in roughly seven weeks, following Argon, Carbon and Hydrogen, and sits between the flagship Carbon and the lightweight Hydrogen. Google describes it as balancing depth and efficiency, and it has taken over Argon's spot in the lineup. The model handles text, spreadsheets, PDFs, charts, audio and video within a single 1M-token context window, supports up to 1M tokens of output, and offers four reasoning-effort levels, with a focus on coding, advanced reasoning and agent tasks.
Reports that followed diverged mainly on availability and pricing. Gemini 4 Neon is available through the Gemini API, Google AI Studio, Android Studio, Google Antigravity, Gemini for business and GitHub Copilot, while in the Gemini app it is limited to Google AI Pro and Ultra subscribers, with Pro priced at $19.99 per month. API pricing is $3 per million input tokens, $0.15 for cached input, and $15 per million output tokens. Artificial Analysis lists the three-tier Gemini 4 pricing as Carbon at $4 and $20, Neon at $3 and $15, and Hydrogen at $1.50 and $7.50 per million input and output tokens, ordered from most to least expensive.
Benchmark and cost claims also conflict. Google says that at the default high reasoning level Neon costs 32% less per task than Carbon while running close to twice as fast. Artificial Analysis measured both models at 60 on the Intelligence Index at high effort, with Neon at $3.74 per task versus Carbon's $7.76, roughly 110 tokens per second, and about 40% shorter task times; at xhigh, Neon scores 61 and rises to $6.44 per task. Cited coding and agent index figures differ, including 58.6 and 57.1 on the Artificial Analysis Coding Agent Index, with another figure at 61. A developer's breakdown noted that Carbon and Argon keep launch discounts of $2 and $10 per million tokens until November 30, making Neon about 50% more expensive over that window; once Carbon moves to $4 and $20 on December 1, Neon becomes roughly 25% cheaper. Current reporting stops at the rollout scope and the discount expiry, with the benchmark and discount discrepancies unresolved.
AI-generated from 20 reports · updated 9 hours ago
Latest turnGoogle shipped Gemini 4 Neon on November 18 with input tokens priced at $3 per million, above the $2 of its Carbon sibling despite sharing the same 1M context window, which sits oddly against the "frontier intelligence at lower cost" pitch. The author argues Neon thinks hard by default, writes more tokens, and finishes tasks faster, so the metric that matters is cost per finished job, not cost per token.
Related AI tools
- Heat index
- 223
- All sources
- 22
- All reports
- 39
- First seen
- 3 days ago
Reports on this story headlines open the original
Google shipped Gemini 4 Neon on November 18 with input tokens priced at $3 per million, above the $2 of its Carbon sibling despite sharing the same 1M context window, which sits oddly against the "frontier intelligence at lower cost" pitch. The author argues Neon thinks hard by default, writes more tokens, and finishes tasks faster, so the metric that matters is cost per finished job, not cost per token.
DEV Community · AIAI score 63
A developer priced out Gemini 4 Neon at $3 per 1M input tokens and $15 per 1M output tokens, with cached input at $0.15 — roughly 50% more than Gemini 4 Carbon's $2/$10 through November 30. By his math, Carbon and Argon double to $4/$20 on December 1, which would make Neon about 25% cheaper, so teams running agent workloads should rerun their cost models before then.
DEV Community · AIAI score 60TestingCatalog reports on X that long-pressing the model selector in Gemini brings up a periodic-table-style menu listing several Gemini 4 models — Hydrogen, Argon, Carbon and Neon — with Nano Banana filed under the Potassium element. The claim comes from a third-party observation of the interface, not a Google announcement.
X:TestingCatalog (@testingcatalog)AI score 55Google released Gemini 4 Hydrogen, a faster and more efficient model built for robotics with stronger spatial and embodied reasoning. It ships with a Robotics plugin for Google Antigravity that adds hardware onboarding, MuJoCo simulation and closed-loop control, which Google says can take a robot from the box to a working application in an afternoon.
Google Developers BlogFirst-partyAI score 69Google DeepMind released Gemini 4 Neon, which matches its flagship Carbon at the default reasoning level and reaches the top three of the Artificial Analysis Intelligence Index at its highest setting. Artificial Analysis measured $3.74 per task at high effort versus $7.76 for Carbon, with output of roughly 110 tokens per second and about 40% shorter task times.
InfoQ 中文AI score 75
According to a Juejin breakdown, Google released Gemini 4 Neon on November 18. It tops an agentic coding leaderboard with 58.6 points, and cached context costs $0.15 per million tokens. The post also includes a master-worker orchestration template.
掘金AI score 60Artificial Analysis reports that Neon and Hydrogen occupy the entire Pareto frontier of intelligence versus cost per task and intelligence versus time per task among three models. The range runs from Hydrogen at low effort (Intelligence Index 27 at $0.29 per task) to Neon at xhigh (61 at $6.44 per task).
X:Artificial Analysis (@artificialanlys)First-partyAI score 44Artificial Analysis rounded up pricing for the three Gemini 4 models: Carbon at $4/$20 per 1M input/output tokens, Neon at $3/$15 and Hydrogen at $1.50/$7.50. Carbon sits at the top of the price range, with Hydrogen the cheapest of the trio.
X:Artificial Analysis (@artificialanlys)First-partyAI score 74Google DeepMind introduced Gemini 4 Neon on 18 November, a mid-tier model positioned between the Carbon flagship and the faster Hydrogen, according to a DEV Community post. The article reports a score of 61 on the Artificial Analysis Intelligence Index, first place on the Coding Agent Index at 57.1, $3 per million input tokens and a one-million-token context window, with rollout across the Gemini API, AI Studio, Antigravity and paid Gemini app tiers.
DEV Community · AIAI score 65
Google has launched Gemini 4 Neon, a mid-tier model positioned between the flagship Carbon and the fast Hydrogen, now generally available in GitHub Copilot. It targets long-running agents and multi-step tool use, though its headline benchmark scores depend on the thinking level used.
DEV Community · AIAI score 64Google has released Gemini 4 Neon, a model that slots between Gemini 4 Hydrogen and Gemini 4 Carbon and replaces Argon in the Gemini lineup. According to The Next Web, it is rolling out now in the Gemini API, Gemini for business, Google Antigravity and the Gemini app, announced in a blog post by Google DeepMind SVP Koray Kavukcuoglu.
The Next WebAI score 57Google’s new AI model Gemini 4 Neon sits between the more capable Carbon and the cheaper Hydrogen, Golem.de reported. The model is pitched as offering a balance of performance and cost.
Golem.deAI score 53According to The Decoder, Google DeepMind has released Gemini 4 Neon, its fourth model in seven weeks. The model is nearly twice as fast as its sibling Carbon, outperforms it on most benchmarks, and comes at a significantly lower price.
The DecoderAI score 79A DEV Community write-up rounds up published Gemini 4 Neon benchmark scores: 58.6 on the Artificial Analysis Coding Agent Index for first place, 65.2% on Terminal-Bench 4.1 and 70.9% on the Vals Index for second. Google also claims the model costs 32% less per task than Gemini 4 Carbon at the default high thinking level. The piece notes most runs used the xhigh level and some came from Google's own harness, and shows how to translate cost-per-task into your own bill.
DEV Community · AIAI score 59Google has released Gemini 4 Neon, a mid-tier model that sits between Carbon for long, hard projects and Hydrogen for fast everyday tasks, which Google describes as balancing depth and efficiency. It is priced at $3 per 1M input tokens, $0.15 per 1M cached input tokens and $15 per 1M output tokens, with a 1M-token context window — pricing that is not aggressive by frontier-model standards.
开源中国AI score 65Google DeepMind released Gemini 4 Neon, a mid-tier model priced at $3 per million input tokens, $0.15 for cached input and $15 for output, with a 1M-token context and four effort levels. It is rolling out to the paid Gemini API, AI Studio, Android Studio and Antigravity, with Pro and Ultra subscribers getting it in the Gemini app first, per Latent Space's AINews roundup.
Latent SpaceAI score 66Google has released Gemini 4 Neon, a 1M-context model priced at $3 per million input tokens and $15 per million output tokens, according to MarkTechPost. It scores 61 on the AA Intelligence Index and ranks first on AA's Coding Agent Index.
MarkTechPostAI score 74Google DeepMind released Gemini 4 Neon on November 18, the fourth elemental model in the Gemini 4 family in seven weeks, according to GeekPark. Official benchmarks put it near or above the flagship on some tasks at roughly half the per-task cost of competitors, with a 1M-token context window, up to 1M tokens of output and native text, audio, video, PDF and spreadsheet input; Google AI Pro subscribers can use it directly in the Gemini app.
极客公园AI score 72According to a Portuguese-language post on DEV Community, Google released Gemini 4 Neon, a model aimed at coding, advanced reasoning and professional tasks. It is available in Google AI Studio, the Gemini API, Android Studio and Google Antigravity, and inside the Gemini app for Google AI Pro and Ultra subscribers. API pricing is $3 per million input tokens and $15 per million output tokens.
DEV Community · AIAI score 58Google has launched Gemini 4 Neon, a model that processes spreadsheets, PDFs, charts and audio-video files within a single 1M-token context window and turns them into structured deliverables. Pricing is $3 per 1M input tokens, $0.15 per 1M cached input tokens and $15 per 1M output tokens.
36氪AI score 77Google has released Gemini 4 Neon, a midtier model that is three times faster than its Gemini 4 Carbon flagship launched a month earlier and beats it on some metrics, SiliconANGLE reported. It sits between Carbon and the cost-focused Gemini 4 Hydrogen.
SiliconANGLEAI score 62Google introduced Gemini 4 Neon on November 18, the fourth variant of the Gemini 4 family after Argon, Carbon and Hydrogen. Google describes it as frontier-level judgment carried to task completion at a cost and latency teams can repeat all day, rather than the smartest model on the roster. The write-up covers its benchmark results, API pricing and how it fits into orchestrator workflows alongside the other three variants.
DEV Community · AIAI score 58Zhidongxi reports that Google has released Gemini 4 Neon, the fourth Gemini 4-series model in seven weeks. Aimed at complex software engineering, knowledge work, and scientific research, it is pitched as faster and more cost-efficient at multi-step tasks at frontier intelligence.
智东西AI score 71Google is rolling out Gemini 4 Neon, priced at $3 per 1M input tokens and $15 per 1M output tokens, 9to5Google reports. The model sits between Hydrogen and Carbon and replaces Argon in the Gemini app, following last week's release of Carbon and Hydrogen.
TechmemeAI score 67Simon Willison released llm-gemini 0.38, adding support for Gemini 4 Neon, which Google introduced today. He argues Neon makes the already confusing Gemini 4 pricing worse: the lineup spans Hydrogen, Argon and Carbon, with Argon and Carbon priced the same and Hydrogen at three-quarters of that price until a December 1 discount expires.
Simon WillisonAI score 52Three weeks after Gemini 4 Carbon, Google released Gemini 4 Neon, a mid-tier flagship that was available the same day across Google AI Studio, the Gemini App, the enterprise API and Google Antigravity, with no waitlist. It is priced at $3 per million input tokens and $15 per million output tokens, ships with a 1M-token context window and scores 61 on the Artificial Analysis index, ifanr reported.
爱范儿AI score 75Google released Gemini 4 Neon, aimed at complex coding, scientific research and enterprise workflows, now available to developers, enterprises and some paid subscribers with a 1M token context window and four reasoning levels from low to xhigh. API pricing is $3 per 1M input tokens and $15 per 1M output tokens, with cached input at $0.15. Google says it ranks first on Artificial Analysis's coding agent index at 58.6 and costs 32% less per task than Gemini 4 Carbon at the default high setting.
IT之家AI score 76Cursor says Gemini 4 Neon is now available in its editor, scoring 63.7% on CursorBench 4.1 at default effort versus 57.3% for Gemini 4 Hydrogen. It is priced at $3 per million input tokens and $15 per million output tokens.
Google on Wednesday released Neon, a new model in the Gemini 4 family. Independent evaluator Artificial Analysis scored it 61 on its Intelligence Index, third globally behind GPT-6.2 Astra and Claude Fable 5.2, with per-task cost down about 17% and completion time roughly 40% shorter than its predecessor Carbon.
华尔街见闻AI score 70Google released Gemini 4 Neon, priced at $3 per 1M input tokens and $15 per 1M output tokens, available in Gemini, AI Studio and Antigravity as well as through APIs. TestingCatalog reports the model ranks #2 on Agent Arena and #3 on the AA Index.
TestingCatalogAI score 71
Other stories people are talking about
- Heat index 77Microsoft rolls out Copilot integrations and unveils MAI models at Ignite 20262 sources
- Heat index 133Google adds GIF generation to Gemini, powered by Omni 1.2 Flash2 sources
- Heat index 33Google Rebrands Spark as Tasks, Shifts to Gemini 4 Hydrogen1 sources
- Heat index 84Google Search and Gemini launch Connected Apps for shopping and tax tasks2 sources
- Heat index 66Google opens Gemini Tasks agent mode to free users in six countries1 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion

