Google has officially launched Gemini 3.7 Flash, and one of the most significant enhancements is the upgrade to Gemini Spark. This updated model is set to revolutionize the capabilities of Google’s cloud-based agent, providing improved tool usage and heightened performance, particularly for multi-step tasks.
Spark’s New Capabilities
Starting today, Gemini Spark will benefit from the advancements in Gemini 3.7 Flash, enabling it to handle tasks such as consolidating files, drafting emails, and updating status documents across Google Workspace efficiently. This upgrade marks a pivotal moment for users, as Spark’s functionality is expanding significantly.
Gemini Spark already operates continuously, even when your laptop is closed or your phone is locked. It can function across various Google services, including Gmail, Drive, Docs, Calendar, Keep, and Tasks. Recently, its capabilities were further enhanced through integration with Chrome, allowing it to utilize saved accounts and passwords for tasks like comparing flight options, scheduling appointments, and initiating bookings. However, users retain control, especially when sensitive actions such as payments are involved.
Improved Workflow and Reliability
Gemini 3.7 Flash is geared towards making everyday workflows more reliable. Google has stated that the model now places greater emphasis on planning and tool calls, demonstrating improved adaptability in the face of obstacles. This means users can expect fewer retries and reduced manual intervention when using Spark.
Benchmark results reflect these positive changes, with Gemini 3.7 Flash achieving 30.4% on AutomationBench, a substantial increase from the 17% scored by Gemini 3.6 Flash. Notably, it outperformed competitors such as GPT-5.6 Terra, which scored 23.6%, and Claude Sonnet 5 at 10.7%.
Strength Beyond Spark
The enhancements provided by Gemini 3.7 Flash are not limited to the Spark upgrade; the model has also demonstrated a significant leap in performance regarding document-heavy tasks. It achieved a score of 34% on GDP.pdf, compared to 22% for its predecessor, 28% for Claude Sonnet 5, and 24.7% for GPT-5.6 Terra.
Google
The pricing structure further amplifies the appeal of these improvements. Under introductory pricing through December 31, Gemini 3.7 Flash is available for $0.75 per million input tokens and $3.75 per million output tokens, making it a more economical choice compared to Claude Sonnet 5 and GPT-5.6 Terra.
Google
Google
Moreover, improvements were also observed in coding-focused benchmarks, with Gemini 3.7 Flash scoring 43.6% on FrontierCode, surpassing Claude Sonnet 5 at 42.7% and GPT-5.6 Terra at 41.3%.
For users of Spark, Gemini 3.7 Flash represents a significant upgrade, particularly if its enhancements in tool usage and multi-step planning translate into reduced task failures and less manual intervention. While these updates are noteworthy, many power users are still eagerly awaiting the ambitious Gemini 3.5 Pro model, which was initially previewed at Google I/O in May but has yet to be released following its anticipated rollout in June.
Image Credit: www.digitaltrends.com
