Google Releases Gemini 4 Carbon, Tops Artificial Analysis Intelligence Index at 59
Google released Gemini 4 Carbon, a frontier AI model arriving two weeks after Gemini 4 Argon, which the company says ranks first on the Artificial Analysis Intelligence Index v5.0 with a score of 59. Google reports leading cybersecurity results, including 88.2% on SEC-bench Pro, 54.8% on ExploitGym and 71.0% on CWE-bench, and cites long-horizon coding gains. Because of its exploitation capability, access begins with a subset of cyber defenders through the Fairwind Program before wider availability. Independent verification remains pending.
IllustrationEditorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page reads the event directly, while its address stays stable when the title changes.
- Summary covers the current reports
Cross-source coverage
Reporting timeline
Google Announces Gemini 4 Carbon, Tops Artificial Analysis V5 Ranking
Google announced Gemini 4 Carbon, a new state-of-the-art model that takes the top spot on the Artificial Analysis V5 ranking, according to a post from testingcatalog. The model is being rolled out to cyber defenders first. According to the post, it shows significant gains across coding and agentic tasks as well as capabilities critical for cybersecurity defense, landing a firm lead in SEC-bench Pro at 88.2% for finding vulnerabilities, ExploitGym at 54.8% for exploitation, and CWE-bench at 71.0% for patching. The post notes that the model still needs testing.
Read sourceGoogle's Gemini 4 Carbon Tops GPT-6 Astra and Claude Opus 5.5 on Artificial Analysis
Google released Gemini 4 Carbon, a new frontier model arriving two weeks after Gemini 4 Argon, which the source says beats GPT-6 Astra and Claude Opus 5.5 on Artificial Analysis. Carbon's largest gains are in security: 88.2% on SEC-bench Pro for finding vulnerabilities, 71.0% on CWE-bench for patching, and 54.8% on ExploitGym for exploiting them. Due to that exploit capability, access begins this week with a subset of cyber defenders, then trusted partners through the Fairwind Program, before developers, businesses, and consumers. Google says Carbon can work on engineering problems for days, root-causing subtle bugs across large codebases. Introductory pricing is $2/$10 per 1M input/output tokens, later doubling to $4 and $20. On the long-context GraphWalks benchmark (256K to 1M tokens, BFS F1), Carbon scores 96.2%, versus Argon's 84.2%, GPT-6 Astra's 71.8%, and Claude Opus 5.5's 66.8%.
Read sourceGoogle's Gemini 4 Carbon Tops Intelligence Index v5, Trails Claude Opus 5.5 on Coding Agents
Google's Gemini 4 Carbon has taken the #1 spot on the Artificial Analysis Intelligence Index v5 with a score of 59, driven by leading performance in complex reasoning, knowledge, and scientific tasks. It ranks first in 4 of 9 Index evaluations, including HLE-Diamond, Terminal-Bench-Science, AA-Omniscience, and AutomationBench-AA. On the Artificial Analysis Coding Agent Index, Gemini 4 Carbon scores 59, second only to Claude Opus 5.5 at 62. The model is being rolled out to selected users and is not yet publicly available, with a 50% launch discount and no confirmed end date. It costs $7.76 per Intelligence Index task, about 23% less than Claude Opus 5.5, and uses 123k output tokens per task. Gemini 4 Carbon has a 1M-token context window, supports text, image, video, and speech input, and is priced at $4/$20 per 1M input/output tokens.
Read sourceShow 5 older updatesHide older updates
Google Unveils Gemini 4 Carbon, Claims Top Spot Over Claude Opus 5.5 and GPT-6 Astra
Google announced Gemini 4 Carbon, which it describes as its most powerful AI model to date. According to the announcement, the model surpasses Anthropic's Claude Opus 5.5 and OpenAI's GPT-6 Astra on the Artificial Analysis Intelligence Index. The post frames the release as a leading benchmark result for Google in the competitive frontier AI model race, though no additional technical details, benchmark scores, or independent verification were provided in the source item.
Read sourceGoogle Announces Gemini 4 Carbon AI Model With Cybersecurity and Long-Horizon Capabilities
Google's AI team announced Gemini 4 Carbon, which it describes as its strongest model to date. According to the post, Carbon features state-of-the-art cybersecurity capabilities and long-horizon capabilities. The announcement was shared via the @GoogleAI account and reposted through the GeminiApp channel. The post provides no further technical details, benchmarks, availability dates, or pricing information beyond these capability claims.
Google Previews Gemini 4 Carbon Frontier Model With SOTA Coding and #1 Intelligence Index Rank
Two weeks after rolling out Gemini 4 Argon, the company previewed Gemini 4 Carbon, described as its next frontier model. According to the post, Carbon has state-of-the-art long horizon coding capabilities that drive SOTA performance across the cybersecurity defense lifecycle, including detection, validation, and remediation. It also ranks #1 overall on the recently updated v5.0 of the Artificial Analysis Intelligence Index with a score of 59. The company said it is beginning rollout this week to a subset of cyber defenders through its Fairwind Program, with access to expand further soon.
Read sourceGoogle Releases Gemini 4 Carbon, Tops Artificial Analysis Intelligence Index at 59
Sundar Pichai announced Gemini 4 Carbon, described as Google's next frontier model, released two weeks after Argon. According to the post, Carbon ranks #1 on the recently updated v5.0 of the Artificial Analysis Intelligence Index with a score of 59. It also leads on cybersecurity evaluations, scoring 88.2% on SEC-bench Pro for finding vulnerabilities, 54.8% on ExploitGym for exploitation, and 71.0% on CWE-bench for patching. Pichai said the model shows significant improvements in long-horizon tasks compared with Argon and that frontier safeguards have been strengthened. Rollout will be gradual and responsible: some trusted cyber defenders gain access starting this week, with expansion to more partners through the Fairwind Program ahead of wider availability.
Read sourceGoogle Launches Gemini 4 Carbon, Tops Artificial Analysis Intelligence Index v5.0
Sundar Pichai announced that Google has released Gemini 4 Carbon, which he said ranks first on the updated Artificial Analysis Intelligence Index v5.0 with a score of 59. The source states that Gemini 4 Carbon posts numbers across multiple benchmarks and has a phased rollout schedule, allowing readers to compare it with Argon and competing products. The announcement is attributed to Pichai via X, and the ranking claim is presented as his statement rather than independently verified in the item.