Google announced its next frontier model, Gemini 4 Carbon, on 2026-10-28: its strongest model yet, with state-of-the-art cybersecurity and long-horizon capabilities. Carbon builds on Gemini 4 Argon, launched about two weeks earlier, ranks #1 on the updated Artificial Analysis Intelligence Index v5.0 at 59, and posts concrete scores across coding, agentic, and cybersecurity-defense benchmarks.
Long-horizon engineering and WebDev Arena
Carbon is built for harder, long-horizon engineering problems and sustained agentic workflows. Internally, Google says teams have seen it run for days, persisting through dead ends on complex multi-step tasks such as root-causing and patching subtle bugs across large codebases and low-level systems. On full-stack web development quality, Carbon jumps +89 Elo to land at #4 (1768) on WebDev Arena.
Cybersecurity defense-lifecycle benchmarks
Google says Carbon achieves SOTA across the cybersecurity defense lifecycle—vulnerability detection, validation, and remediation. External benches (evaluated without guardrails, per the post): SEC-bench Pro 88.2% (finding real vulnerabilities), ExploitGym 54.8% (validating exploitability), and CWE-bench 71.0% (root-cause patching). Internally: more than 2× uplift vs Argon in success rate for generating secure code; the Chrome team reports more than 10× better high-quality patch generation than another lab’s large frontier model; cloud-security platform Wiz observes Carbon leading on web penetration-testing benchmarks (tied to its Scan for Good initiative). This brief reports official bench names and scores only—no exploit steps, PoCs, or attack procedures.
Alignment and prompt-injection resistance
Carbon keeps Argon’s multi-layered safety stack (training, real-time misuse defenses, prompt-injection resilience, chain-of-thought monitoring) and pushes alignment further. Official figures: successful cheating on Terminal-Bench 4.0 cut to 0.0% (cheat attempts 0.4%); high-severity misalignment alerts in live internal coding environments down almost 80%; unsupported claims cut by about 2× vs Argon; overreach −26% and under-service −29%; Gray Swan Indirect Prompt Injection (IPI) attack success 0.0% (Argon previously 0.7%).
Fairwind, government pre-release testing, and pricing
Rollout is deliberate: starting this week, Carbon goes first to a subset of cyber defenders, then gradually to additional trusted partners via the Fairwind Program; Google is also in the U.S. government’s voluntary pre-release model testing process. Availability to developers, businesses, and consumers on product surfaces is “as soon as possible”—no full consumer/API GA date is stated. Introductory price matches Argon: $2 per million input tokens and $10 per million output tokens, with cached input at 95% off input price; after the introductory period, $4 / $20.
Primary sources:
- Google, “Introducing Gemini 4 Carbon” / “Gemini 4 Carbon: our next frontier model,” 2026-10-28, https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/
