Google releases Gemini 3.5 Flash globally, claims it's 'strongest agentic and coding model yet'
Google released Gemini 3.5 Flash at its I/O 2026 conference, immediately deploying it as the default model in AI Mode and the Gemini app globally. The company claims it delivers improved agentic capabilities and coding performance at less than half the cost of competing frontier models. Gemini 3.5 Pro is scheduled for release in June.
Gemini 3.5 Flash at a glance
How Gemini 3.5 Flash ranks now
| Benchmark | Score | Rank | Source |
|---|---|---|---|
| Arena Elo | 1478 Elo | 26 of 141 | LMArena, Nov 2026 ↗ |
| GPQA | 92.2% | 25 of 164 | Artificial Analysis, Jul 2026 ↗ |
| Speed (tok/s) | 167 tok/s | 17 of 103 | Artificial Analysis, Nov 2026 ↗ |
Rank is against every model in our database with a sourced score on that benchmark, each counted once with its best score.
Google releases Gemini 3.5 Flash globally, claims it's 'strongest agentic and coding model yet'
Google released Gemini 3.5 Flash at its I/O 2026 developer conference on May 19, immediately deploying it as the default model powering AI Mode in Search and the Gemini app globally. The company claims the model represents its "strongest agentic and coding model yet," though specific benchmark scores were not disclosed.
Model release details
Gemini 3.5 Flash is the first model in Google's 3.5 series, with Gemini 3.5 Pro scheduled for release in June 2026. According to Google, the model can help developers complete tasks "in a fraction of the time, often at less than half the cost of other frontier models," though the company did not provide specific pricing per million tokens.
The model is now available in AI Mode in Search, the Gemini app, and various developer tools including Google Antigravity. Google says data scientists have been using the model to extract insights from complex data.
Safety improvements
Google claims Gemini 3.5 is less likely to generate harmful material or incorrectly refuse legitimate queries. The company says it employs "more advanced safety measures," including checks on the model's reasoning before generating responses. No specific benchmarks for these safety improvements were provided.
Search overhaul and AI agents
Google announced a comprehensive redesign of its Search box, calling it "the biggest upgrade to our Search box in 25 years." The new intelligent Search box accepts text, images, files, videos, and Chrome tabs as inputs, and is rolling out in all languages and countries where AI Mode is available.
The company is introducing "information agents" that can monitor blogs, news sites, social media, real estate listings, and real-time data on finance, sports, and shopping. These agents will be available to Google AI Pro and Ultra subscribers this summer.
Additionally, Google is deploying agentic booking features in Search for all US users this summer. The system can explore availability and pricing for services, and in some categories including beauty, pet care, and home repair, can call local businesses on behalf of users.
Generative UI and coding tools
Google is integrating its Antigravity tool into Search to enable users to generate custom visualizations, tables, graphs, and simulations at no charge. The feature will be available to all users this summer.
AI Pro and Ultra users will also gain access to generate mini apps for tasks like wedding planning and fitness tracking, with Antigravity able to pull in Google data including weather, maps, and reviews.
Personal Intelligence expansion
Google expanded its Personal Intelligence feature, which allows Gemini to access information from Gmail, Google Photos, and soon Google Calendar, to 98 languages across nearly 200 countries and territories. The feature was previously made free for US users in March 2026.
What this means
Google's aggressive push of AI features into Search represents a fundamental shift in how the company expects users to interact with web information. The move to make Gemini 3.5 Flash the default model in AI Mode globally signals confidence in the model's capabilities, though the lack of disclosed benchmarks or pricing makes direct comparisons with competing models difficult. The expansion of AI agents and generative UI tools positions Google to compete directly with specialized AI assistants, though the impact on web publishers who have already seen traffic declines from AI Overviews remains a concern.
Related Articles
Google releases Genie 4 to AI Ultra subscribers with real-time prompt editing and native audio
Google is rolling out Genie 4, which it calls its most realistic and interactive world model yet, to eligible Google AI Ultra subscribers in Project Genie. The update adds real-time native audio, mid-session text prompting, a world-building agent and a community gallery. Genie 3 also gains sessions of up to 3 minutes.
RadixArk releases DaisyRose-27B, lifting Terminal-Bench 4.1 pass@1 from 5.6% to 12.0%
RadixArk released DaisyRose-27B, its first model, built on Qwen3.8-27B and trained with reinforcement learning on 1,200 curated agent tasks. According to RadixArk, it raises Terminal Bench 4.1 pass@1 from 5.6% to 12.0% and scores higher than the base model on all seven reported benchmarks. Weights ship under Apache 2.0 in BF16, FP8 and NVFP4.
Microsoft releases Aion 1.0 Instruct, a 2.3B on-device SLM with 32K context under MIT license
Microsoft's Windows Applied Sciences Group has published Aion 1.0 Instruct, a roughly 2.29B-parameter instruction-tuned small language model with a 32,768-token context window under the MIT license. A further optimized version is slated to become the Windows OS SLM by the end of January 2027.
Bria releases RMBG-3.0, a 4-step background remover built on Fibo 2.0, with open weights for non-commercial use
Bria AI has released RMBG-3.0, the third generation of its background removal models and the first built on a generative foundation. It is Fibo 2.0, fine-tuned and distilled to produce an alpha matte in 4 steps. Bria claims it ranks first among six systems in its own benchmark of 234 hard images, with an average rank of 1.51.
Comments
Loading...