Google has introduced the discharge of Gemini 3 Flash, its newest frontier mannequin designed for pace at a decrease token value.
In response to Google, this mannequin is good for iterative improvement, because it is ready to rapidly cause and clear up duties in high-frequency workflows. It additionally outperforms all Gemini 2.5 fashions in addition to Gemini 3 Professional in coding capabilities on SWE-bench Verified.
Moreover, as a consequence of its sturdy efficiency in reasoning, software use, and multimodal capabilities, it’s very best for duties like complicated video evaluation, information extraction, and visible Q&A, enabling extra clever functions that demand superior reasoning and fast solutions, like in-game assistants or A/B check experiments.
Gemini 3 Flash is now out there for builders within the Gemini API by way of Google AI Studio, Gemini CLI, and Google Antigravity. It’s additionally out there within the Gemini app, AI Mode in Search, Vertex AI, and Gemini Enterprise.
“With Gemini 3, we launched frontier efficiency throughout complicated reasoning, multimodal and imaginative and prescient understanding and agentic and vibe coding duties. Gemini 3 Flash retains this basis, combining Gemini 3’s Professional-grade reasoning with Flash-level latency, effectivity and value. It not solely allows on a regular basis duties with improved reasoning, but in addition is our most spectacular mannequin for agentic workflows,” Tulsee Doshi, senior director of product administration at Google, wrote in a weblog publish.
