With Gemini 4 Neon scoring 61 on the Artificial Analysis Index at $3.00 input and $15.00 output per million tokens, Alphabet undercuts its own flagship Carbon tier by 25.0% while collapsing four releases into a unified three-tier software stack.
56.51·NeutralResearchPublished Nov 19, 2026 · Data as of Nov 18, 2026· 2 min read · 1 exhibitUpdated
Executive thesis and the Gemini 4 product ladder
Neon quality score
61
AA Index v5.0.1
Leads Carbon (60) y/y
Neon flat 1M price
$3.00 / $15.00
Per 1M in / out
-25.0% vs. Carbon y/y
Hydrogen flat 1M price
$1.50 / $7.50
Per 1M in / out
5/5 speed tier y/y
Production coding lead
39%
SWE-sweep harness
3.86x nearest peer y/y
Most superintelligence model releases force software engineering teams to choose between frontier accuracy and usable latency. Alphabet's Gemini 4 lineup breaks that trade-off by separating reasoning depth, execution speed, and distribution surfaces into explicit elemental tiers [1][3].
Over a three-week stretch ending November 18, 2026, Google DeepMind rolled out Gemini 4 Carbon, Gemini 4 Argon, Gemini 4 Hydrogen, and finally Gemini 4 Neon, each sharing a unified 1M-token context window with flat token pricing and configurable thinking budgets [1][3].
The strategic pivot inside this launch is the price-quality inversion created by Neon. On Google's published November 18 lineup card, Neon posts a 61 on the Artificial Analysis Index v5.0.1 compared to 60 for Carbon and 54 for Hydrogen, yet Neon runs noticeably faster (3.5 out of 5 speed rating vs. Carbon's 2 out of 5) and lists at $3.00 input / $15.00 output per million tokens versus Carbon's standard $4.00 / $20.00 rate [1][2].
Argon, which initially launched alongside Carbon as a heavy software engineering and analysis workhorse, is effectively superseded by Neon across both developer tooling and consumer tiers.
Exhibit 1Financial Schedule
Exhibit 1: Gemini 4 Four-Model Architecture, Flat 1M Pricing, and Distribution Map
Official November 18, 2026 Gemini 4 tier positioning across quality, flat 1M-token API pricing, thinking controls, speed, and Alphabet surfaces.
Model Tier
Primary Software Role
AA Index
Flat 1M Price (In / Out)
Thinking Levels
Speed
Alphabet Surfaces
Gemini 4 CarbonIntroductory $2.00 / $10.00 promotional pricing runs through November 30, 2026.
Hardest long-horizon architecture and deep research
60
$4.00 / $20.00
Low, medium, high
2.0 / 5
Gemini API Studio (paid), Gemini for Business, Ultra app, Antigravity, Search mode
Gemini 4 ArgonCollapsed into Neon to simplify developer model selection without sacrificing coding depth.
Initial heavy code and enterprise analysis tier
Superseded
$4.00 / $20.00
Low, medium, high
Legacy
Rolled out alongside Carbon; absorbed into Neon's faster, lower-cost tier
Gemini 4 Neon25.0% cheaper than Carbon's list price while adding an xhigh thinking budget.
Deep reasoning across coding, knowledge work, and science
61
$3.00 / $15.00
Low, medium, high, xhigh
3.5 / 5
Gemini API Studio (paid), Gemini for Business, Pro & Ultra app, Antigravity (paid)
Gemini 4 HydrogenAvailable to everyone on consumer tiers; serves as the high-speed worker in routed workflows.
Speed and efficiency in everyday coding and agentic execution
54
$1.50 / $7.50
Low, medium, high, xhigh
5.0 / 5
Gemini API Studio (all tiers), Gemini for Business, Gemini app, Google Search, Antigravity
Source: Google Blog [1], DeepMind Evaluation Methodology [2], and Gemini API Documentation [3].
Sector Peer
Full Access
Dynatrace: Buy on Profitable Enterprise Observability at a 32% Discount to Base DCF
Dynatrace combines 21.4% TTM revenue growth and 81.1% gross margin with positive 12.7% GAAP operating margin and $578.7M of GAAP FCF, trading at $42.63 versus our $62.81 5-year DCF base value.
Dynatrace, Inc.Nov 8, 2026 · Data as of Sep 30, 2026