Model

Google's Gemini 4 Neon took the coding lead the day before Anthropic's $1.5 trillion debut

Google's mid-tier model debuted at No. 1 on Artificial Analysis's coding-agent index at $6.86 a task against Claude Opus 5.5's $12.61, a day before Anthropic's Nasdaq debut. A $1.5 trillion valuation built on coding-led growth now meets a measured, cheaper rival, and the exposure sits with whoever paid $83.

Richard TangRichard Tangthree-minute read
Share
Portrait of Anthropic co-founder and chief executive Dario Amodei, wearing glasses and a navy blazer
Dario Amodei, co-founder and chief executive of Anthropic, which priced its IPO at $83 a share on November 18 and began trading on the Nasdaq the next morning.

Google DeepMind made Gemini 4 Neon generally available on November 18, and the mid-tier reasoning model debuted at No 13. 1 on the coding-agent index run by Artificial Analysis, an independent benchmarking service, at $6.86 per task against $12.61 for Claude Opus 5.5, Anthropic's$350B — Anthropic, private, latest valuation $350B flagship 13. The same day, Anthropic priced its IPO at $83 a share for a fully diluted valuation of about $1.5 trillion, and the stock trades on the Nasdaq today under the ticker ANT 2. CNBC attributes that growth largely to Claude Code, Anthropic's coding tool, the exact market where the cheaper No. 1 just landed 2.

The rank and the price moved together

Neon is adapted from Gemini 4 Carbon, the larger model in Google's Gemini 4 family, and lists at $3 per million input tokens and $15 per million output, against Opus 5.5's $4 and $20 3. Artificial Analysis put Neon one point clear of both Anthropic flagships, Opus 5.5 and Claude Fable 5.2 1. The measured saving is list price, not token efficiency: Neon burned more output tokens per task than Carbon 1.

Google's new coding leader undercuts Anthropic's flagships on cost per task

The No. 1 coding agent costs about half Anthropic's flagships per task

$0$5$10$15$20Gemini 4 NeonClaude Opus 5.5Claude Fable 5.2$15.47$6.86
Data
Cost per coding task
Gemini 4 Neon$6.86
Claude Opus 5.5$12.61
Claude Fable 5.2$15.47
Average cost per task on the Artificial Analysis Coding Agent Index, measured by Artificial Analysis in November 2026.1

Why the undercut lands on Anthropic

The asymmetry is structural, and builders named it before the index moved. A much-shared Hacker News comment argues that Google−2.92% — Google, down 2.92 percent today can price inference without margin because DeepMind is subsidized by the rest of Google's empire, while for Anthropic and OpenAI$1.18T — OpenAI, private, latest valuation $1.18T "selling inference is their entire business model" 4.

Anthropic has committed $900 billion, mostly to compute, against a rival that does not need margin on inference 5. Near-parity at half the cost per task is likely enough to cap what enterprises will pay for coding agents; no source yet measures the effect on Claude Code revenue or margins, so that cap is price logic, not reported data.

Alphabet holds both sides of the trade

Google owns about 13.7% of Anthropic, worth about $200 billion at the IPO price, and it now sells the cheaper challenger through its own API 63. Seven of Anthropic's suppliers and partners placed $26 billion at the IPO price 6. With the exception of TSMC−3.20% — TSMC, down 3.20 percent today, TechCrunch notes, Anthropic has partnerships with those companies that would see it purchasing their products 5. Gil Luria, head of technology research at D.A. Davidson, set the stake plainly: if everyone ends up using other models, Anthropic is "worth nothing" 5.

One index point is not a lost frontier

The caveats run both ways. Artificial Analysis's coding index is the only independent measurement in this story; the rest of Google's launch chart is self-reported, and it still shows Opus ahead on Terminal-bench 4.1 and FrontierCode 2.0 3. Neon does not need to win to change the price. For builders this is a price cut, and cheaper intelligence is how AI reaches more work; the exposure sits with whoever paid the debut price believing the frontier lead holds.

What to watch now is Anthropic's first public disclosure of Claude Code pricing or revenue, and the next Artificial Analysis index update. At the Nasdaq bell on Thursday, Anthropic's chief executive Dario Amodei recommitted to pacing AI development, taking "as long as is necessary" to keep AI systems under human control 7. If Opus reclaims the top spot, the $83 premium reads one way; if the gap widens, the other 7.

Get Model by email, free

The AI labs, their models, and the business behind them. In our daily email, every weekday.

Loading Google DeepMind

Reader comments

0 comments

    Model

    ElevenLabs, valued at $22 billion, puts its voice agents inside NICE, Genesys, Amazon Connect and Five9

    Salesforce and Five9 have put more than $3 billion into buying customer-service AI this month. ElevenLabs, valued at $22 billion after a September share sale, embedded its agents natively into Five9's platform and three rivals instead, and the evidence points to pricing power sitting with the channel, not the model maker.

    Model

    OpenAI's agent bill now runs at twice its researchers' cash pay

    OpenAI's own report prices agent use at twice its researchers' cash pay while internal deployment is 2 percent of research compute, and the number lands the week Anthropic lists and OpenAI settles for half its raise. The bill stacks inference on the chip budget, and holders pay it.

    Model

    Anthropic's Claude and OpenAI's GPT-6.2 each top a 3D leaderboard during Anthropic's IPO week

    Anthropic, which is expected to price its Nasdaq IPO tonight, holds the crown on one human-vote 3D leaderboard and OpenAI holds the other. Spline, the vendor running the first board, pays its voters in its own AI credits and may hand their votes to research partners.

    Model

    HeyGen's Avatar 6 scored 84% on realism, but its 42.2% pick rate was below chance

    Avatar 6 launched November 17, 2026, with an 84% realism figure and first place in a blind arena; the score doubles a below-chance pick rate, and the arena capped 97% of clips at 35 seconds. That is the evidence base under a company claiming $200 million of annual revenue.

    Model

    MiniMax hit a record low, about 90% below its January peak, as a US rule gates its rented Nvidia compute

    Washington's new remote-access rule puts a US license between China's listed model makers and the offshore Nvidia capacity they rent. MiniMax, the purest listed bet on Chinese foundation models, holds the price tag.

    Model

    OpenAI cuts flagship pricing 20% to match Anthropic as heavy users call GPT-6.2 Astra unusable

    OpenAI cut its flagship's price to match Claude Fable 5.2's rate, and the users who run models hardest rejected it, 24 hours before Anthropic priced a $1.4 trillion IPO on the assumption the quality gap holds. The independent labs say the gap is real, for now.

    Subscribe to TheInference for free