Gemini 3.8 model releases

Google DeepMind's fast, cost-efficient Flash model released Sep 2, 2026 — its fourth Flash model in under four months and the successor to Gemini 3.7 Flash. At high reasoning it scores 59 on the Artificial Analysis Intelligence Index (up 3 points from Gemini 3.7 Flash and level with sub-maximum efforts of GPT-5.6 Sol and Grok 4.6), 57 at medium (matching GPT-5.6 Terra and Muse Spark 1.2), and 52 at low (matching Gemini 3.6 Flash at ~30% lower cost per task). The improvement is driven mainly by agentic evaluations — t^3-Banking tool use (+12 points to 45%), Terminal-Bench v2.1 coding, and GDPval-AA v2 real-world tasks. It keeps a 1M-token context window and multimodal input (text, image, video, speech) with text output. Pricing matches Gemini 3.7 Flash's current discounted rate of $0.75/$3.75 per Mtok input/output through the end of 2026 ($1.50/$7.50 at standard pricing), with cached input keeping a 90% discount; a ~30% rise in average output tokens per task (to ~48k) lifts cost per task to ~$0.58 at high reasoning despite unchanged per-token pricing. Available in the Gemini app for AI Pro and Ultra subscribers, AI Mode, and Gemini in Google Sheets, and for developers via Google Antigravity, AI Studio, and the Gemini API. Benchmark figures are vendor/third-party-reported.

The cybersecurity-tuned sibling of Gemini 3.8 Flash, introduced by Google DeepMind on Sep 2, 2026 in the same launch. It shares the same foundational intelligence as Gemini 3.8 Flash but ships with a more permissive set of mitigations for defensive security work, and is available only to trusted defenders — government authorities, critical-infrastructure operators, and software maintainers — through the new Fairwind Program. Google reports frontier-level autonomous vulnerability discovery (surpassing Gemini 3.5 Flash Cyber and much larger frontier models on the CyberGym benchmark), a success rate exceeding 70% on an internal real-world vulnerability benchmark spanning 20 programming languages, and 47.2% pass@1 on the external CWE-Bench patching benchmark (on the Pareto frontier versus a leading frontier model at 47.8%, at far lower cost). Google is already using it internally: the Chrome Security team reports 2.6x more correct patches than the best larger commercial models, and its Cloud Vulnerability Research team found a critical foundational vulnerability in under two hours. Keeps a 1M-token context and multimodal input with text output; prioritizes vulnerability fixing over offensive capability. Not publicly token-billed (trusted-access only), so no list price is recorded. Benchmark figures are vendor-reported.