✉ The Friday AI Brief: the week's 5 best AI stories, tools & comparisons — in your inbox every Friday morning.

Google logo

Google Put Anthropic’s Claude 5.5 Models Inside Its Own Coding Tool — But Only Paying Users Get Them

Google Put Anthropic’s Claude 5.5 Models Inside Its Own Coding Tool — But Only Paying Users Get Them

October 3, 2026. Google quietly updated Antigravity, its agentic AI coding workspace, to offer Anthropic’s two newest models — Claude Opus 5.5 and Claude Sonnet 5.5 — directly in the model picker. The catch: they’re reserved for paying Google AI Pro and Ultra subscribers. Free users, Google AI Plus members, trial-Pro accounts, and even Enterprise customers are left out.

What happened

On October 3, Google’s Antigravity model documentation began listing Claude Sonnet 5.5 (thinking) and Claude Opus 5.5 (thinking) as available reasoning models for eligible subscribers. Antigravity is Google’s autonomous coding environment — it plans, writes, and tests code across large software repositories on its own.

With the change, developers can now run Anthropic’s flagship models inside Google’s tool without needing a separate Anthropic API key. That matters because Claude has become many developers’ preferred coding model, even in an IDE built by Google.

The details

According to the model documentation, the access matrix is narrower than a simple paid-versus-free split:

  • Claude Opus 5.5 and Sonnet 5.5: Google AI Pro (full, non-trial subscriptions) and Google AI Ultra only.
  • Excluded: Free accounts, Google AI Plus, promotional or trial Pro entitlements, and Enterprise accounts.
  • Retiring November 2, 2026: Claude Sonnet 4.6, Claude Opus 4.6, and the open-weights GPT-OSS-120B are all marked for removal from Antigravity on that date. The Claude 4.6 models currently remain available to Free, Plus, and Pro users — so some users will actually lose models they’re using today.
  • The docs also list Gemini 3.8 Flash, Gemini 3.7 Flash, Gemini 3.6 Flash, and Gemini 3.1 Pro across all displayed plan categories, so Google’s own models stay widely available.
  • One quirk worth knowing: a reasoning-model selection persists within a conversation. Switching models while an agent is mid-run doesn’t change the model being used for that active turn.

Opus 5.5 is positioned as the heavyweight for deep, long-horizon coding and agentic workflows; Sonnet 5.5 is the faster, cheaper option for everyday coding tasks.

Why it matters

This is Google selling its rival’s product inside its own flagship developer tool — which tells you everything about how the model race actually works now. Google knows many developers will reach for Claude over Gemini for coding regardless of whose IDE they’re in, so rather than fight that preference, it bundled the models and put them behind its paid tiers.

For subscribers, Antigravity becomes a genuine multi-model coding environment: Gemini for Google’s ecosystem work, Claude for the hard coding problems, all in one place. For everyone else, it’s a reminder that the best models keep consolidating behind paywalls — and that “free” tiers increasingly mean “last year’s models.”

Team leads comparing Antigravity runs should note the eligibility matrix too: a result produced with a Claude 5.5 entitlement on a Pro account may not be reproducible on a Free, Plus, trial-Pro, or Enterprise account. Benchmarks without subscription context are apples and oranges now.

FAQ

What is Google Antigravity?
Antigravity is Google’s agentic coding workspace — an autonomous development environment that plans, writes, and tests code across complex codebases, with a picker for choosing which AI model powers the agent.

Who gets access to Claude 5.5 in Antigravity?
Only paying Google AI Pro (non-trial) and Google AI Ultra subscribers. Free, Google AI Plus, trial-Pro, and Enterprise accounts are excluded.

What’s the difference between Claude Opus 5.5 and Sonnet 5.5?
Opus 5.5 is Anthropic’s flagship model, aimed at deep, complex, long-horizon coding work; Sonnet 5.5 is the faster, more affordable option for everyday coding tasks.

Are older models being removed?
Yes. Claude Sonnet 4.6, Claude Opus 4.6, and GPT-OSS-120B are scheduled for removal from Antigravity on November 2, 2026 — a real downgrade for Free and Plus users who currently have access to the 4.6 models.

Do I need an Anthropic API key to use Claude in Antigravity?
No — the models run through Google’s integration, which is part of the point. You’re paying Google, not Anthropic.

Sources: Let’s Data Science, PC-Tablet, Startup Fortune, ExplainX.

Leave a Comment

Your email address will not be published. Required fields are marked *

Get the 5 best AI tools every week

Top AI news, tools, and prompts — one short email. Free, unsubscribe anytime.

Scroll to Top