Google Gemini 2.5 Flash Reduces Enterprise AI Agent Token Costs

Google has released Gemini 3.6 Flash and Gemini 3.5 Flash-Lite as next-generation workhorses engineered to dramatically cut latency and token costs for enterprise AI agents operating at scale.

The economics of running autonomous software agents inside a production environment come down to a fixed equation few vendors advertise directly. A model needs to reason through multi-step tasks competently — but every extra token it generates adds cost and delay to workflows that may run thousands of times per hour.

Teams building background agents rather than chat interfaces need throughput first and parameter count second. Google's answer, announced this week, splits that trade-off across three distinct models:

• 🔵 Gemini 3.6 Flash — optimized for coding and multimodal reasoning
• ⚡ Gemini 3.5 Flash-Lite — built for high-volume, low-latency workloads
• 🔒 Gemini 3.5 Flash Cyber — a restricted variant purpose-built for vulnerability remediation

📊 The Math Behind Gemini 3.6 Flash

Google's developer documentation for 3.6 Flash centres on one headline figure: 17% fewer output tokens than the prior 3.5 Flash version, based on measurements from the Artificial Analysis Index.

In specific synthetic tests — including the Datacurve DeepSWE benchmark — Google reports drops in token usage of up to 65%, with pricing set at $1.50 / 1M input tokens and $7.50 / 1M output tokens.

Key benchmark results for Gemini 3.6 Flash vs. its predecessor:

🏢 Figma, Hebbia, and Harvey Put the Model to Work

Figma has integrated Gemini 3.6 Flash into its prototyping infrastructure. According to Matt Colyer, Director of Product Engineering at Figma, the model gives developers a faster route through design iterations without any drop in output quality.

Legal technology platform Harvey and research tool Hebbia route data through the model for multimodal document work — ingesting raw financial filings, parsing document structure, reading embedded charts, and producing draft reports ready for human review.

Google also folded a client-side computer-use tool directly into the Gemini API and Gemini Enterprise platforms — removing the custom intermediary software engineers previously had to build to let models operate on top of an operating system.

The company reports an OSWorld-Verified score of 83.0%, up from 78.4%, and says updated safeguards against chemical, biological, radiological, and nuclear misuse improve resistance to jailbreaking — without raising refusal rates for legitimate requests.

⚡ A Cheaper Tier for High-Volume Background Agents

Gemini 3.5 Flash-Lite targets a different job: document processing and agentic search running at volume rather than reasoning depth. The Artificial Analysis Index measured the model at 350 output tokens per second — the fastest in the 3.5 series according to Google.

Pricing runs at just $0.30 / 1M input tokens and $2.50 / 1M output tokens — cheap enough that engineering teams can route simple, high-volume subagent requests to a minimal thinking level and reserve higher thinking levels for complex multi-step work.

On Google's GDM-MRCR v2 long-context test, Gemini 3.5 Flash-Lite recorded a 72.2% success rate against 60.1% for its predecessor. Its GDPval-AA v2 score nearly doubled — jumping from 642 to 1140. The model carries the same native computer-use tool as 3.6 Flash.

Separately, Google confirmed that Gemini 3.5 Pro remains in partner testing ahead of a full public release, and pre-training for the next Gemini 4 architecture is already underway.

🔒 Gemini 3.5 Flash Cyber: A Restricted Model for Patching Code

Automated vulnerability scanners now surface flaws faster than most security teams can patch them — and that growing gap is precisely where Google positions Gemini 3.5 Flash Cyber.

The model is built to validate and remediate code vulnerabilities, with Google reporting performance on the CyberGym benchmark competitive with frontier models — though detailed figures have not been made public in the same way as consumer-facing releases.

⚠️ Distribution stays restricted to governments and vetted partners through a pilot programme — a limitation Google frames as a safeguard against the model generating exploit code for offensive use.

Inside Google's CodeMender security agent, multiple instances of 3.5 Flash Cyber run in parallel, cross-checking one another's findings before producing a single remediation report that a human reviewer signs off on.

🚀 How to Access These New Models

Engineering teams seeking to integrate these new models can access them through the following channels:

• 🔗 Gemini API via Google AI Studio
• 📱 Android Studio
• 🏢 Gemini Enterprise Agent Platform

Consumers can also access the new models directly in the Gemini app, and Gemini 3.5 Flash-Lite is additionally rolling out within Google Search — signalling that this efficiency-focused tier is set to become a foundational layer across Google's entire product ecosystem.