Google Introduces Gemini 3.6 Flash, 3.5 Flash-Lite and 3.5 Flash Cyber

Google announced Gemini 3.6 Flash, Gemini 3.5 Flash-Lite and Gemini 3.5 Flash Cyber as new additions to its Gemini model lineup.

Google expands the Gemini Flash family

In a Google Blog post, the company said Gemini 3.6 Flash is designed to improve performance across coding, knowledge work and multimodal tasks, while also improving token efficiency and output-token pricing. Google positioned the model as part of its Flash family, which is intended for faster and more cost-conscious AI workloads compared with larger flagship models.

Google’s AI for Developers documentation lists gemini-3.6-flash as a generally available production model. The same documentation says the model supports a 1 million-token context window and up to 64,000 output tokens, and includes pricing and migration guidance for developers using the Gemini API.

Flash-Lite targets lower-cost production use

Google also introduced Gemini 3.5 Flash-Lite. According to Google’s developer documentation, gemini-3.5-flash-lite is also listed as a generally available production model, with a 1 million-token context window and a maximum output length of 64,000 tokens.

The Google Blog describes Flash-Lite as part of the company’s effort to offer more efficient models for developers who need lower latency or lower cost. Cinco Días, the Spanish business publication owned by El País, also reported that Google launched Gemini 3.6 Flash, 3.5 Flash-Lite and 3.5 Flash Cyber, and said Flash and Flash-Lite are available through Google’s developer platforms and apps.

A security-focused Gemini model

The third model, Gemini 3.5 Flash Cyber, is presented by Google as a more specialized model for cybersecurity-related work. The available source material identifies it as part of the same launch, but gives fewer implementation details than the developer documentation provides for Gemini 3.6 Flash and Gemini 3.5 Flash-Lite.

Because the cited developer documentation specifically lists Gemini 3.6 Flash and Gemini 3.5 Flash-Lite as production models, developers evaluating availability should consult Google’s model documentation for the exact model IDs, supported features and pricing before migrating workloads.

Independent benchmarking points to faster task completion

Artificial Analysis said it benchmarked Gemini 3.6 Flash and Gemini 3.5 Flash-Lite ahead of release. According to the research firm, both models roughly halved time per task compared with their predecessors while improving token efficiency.

Those findings are separate from Google’s own launch claims and should be read as one benchmark provider’s measurements rather than a universal result. Actual speed, cost and quality will depend on workload type, prompt design, output length and deployment settings.

What it means for developers

The most concrete change for developers is the addition of new production model options in Google’s Gemini API documentation. Gemini 3.6 Flash appears aimed at teams that need a general-purpose model with improved performance and efficiency, while Gemini 3.5 Flash-Lite is positioned for applications where cost and speed are especially important.

Google’s launch also shows continued segmentation of the Gemini lineup, with general-purpose, lower-cost and specialized cybersecurity variants. For organizations already using Gemini models, the practical next step is to compare the listed model IDs, pricing and migration notes in Google AI for Developers documentation against existing latency, quality and budget requirements.