Crypto
Home›Crypto›Market Structure›Google launches Gemini 3.6 Flash and 3.5 Flash-Lite
Google launches Gemini 3.6 Flash and 3.5 Flash-Lite
Gemini 3.6 Flash cuts output token costs to $7.50 per million and uses 17% fewer output tokens than 3.5 Flash, targeting cheaper AI agent runs at scale.
Google has launched Gemini 3.6 Flash and Gemini 3.5 Flash-Lite, positioning the company’s newest releases as more efficient and lower cost options than the prior Gemini 3.5 Flash line. Decrypt reports the change comes as Gemini 3.5 Pro, originally promised for June delivery and expected from Google I/O 2026, remains in testing after missing internal targets, particularly on coding tasks.
The new models are aimed at AI agents, software that can handle tasks semi autonomously such as managing documents, processing data pipelines, or browsing without a human step by step workflow. Decrypt notes the Flash series is speed optimized, while Pro models are designed as heavier, slower systems for complex reasoning.
Google says Gemini 3.6 Flash uses 17% fewer output tokens than 3.5 Flash, and it lowers output pricing from $9 to $7.50 per million tokens. Decrypt also reports the input price for 3.6 Flash is $1.50 per million tokens.
On performance benchmarks cited by Decrypt, Gemini 3.6 Flash scored 49% on DeepSWE v1.1 versus 37% for 3.5 Flash, and it posted 63.9% on MLE-Bench versus 49.7% for 3.5 Flash. Decrypt also says Google has begun pre training for Gemini 4, describing it as its most ambitious pre training run yet, while the Pro delay follows earlier missed deadlines that contributed to a roughly 4.4% drop in Alphabet stock on the report.