Google Releases Gemini 3.5 Flash (Low) to Fix Token Issues on Antigravity
In Focus
- New model variant optimizes token usage for simple tasks
- Gemini 3.5 Flash (Low) generates 45% fewer tokens than 3.5 Flash (Medium)
- Google has also updated Gemini token quotas across free and paid plans
Google has released Gemini 3.5 Flash (Low) after users raised concerns about tight limits in Antigravity. The latest move is expected to optimize token usage for simple tasks. The newly introduced Low version generates about 45% less tokens compared to the original model.
“We heard concerns that Antigravity consumes many tokens for simple tasks now. So, we’re adding Gemini 3.5 Flash (Low) as a way to optimize token usage for these tasks. In our internal testing, it generates around 45% fewer tokens than Gemini 3.5 Flash (Medium) and generally outperforms Gemini 3 Flash (High) on software engineering tasks,” Google DeepMind Director, Varun Mohan stated in a post on X.
What Issue Will Gemini 3.5 Flash Low Fix?
Gemini 3.5 Flash (Low) is designed to fix Google’s Gemini token quota issues. The search giant has been struggling to provide sufficient tokens to users of its agentic coding platform, Antigravity. Last week, Google announced that it was shifting from the daily prompts tallying pricing system to compute-based usage.
The compute-based system analyses usage based on AI features used, complexity of individual requests, and the length of chat history. This shift meant that intense coding consumes huge volumes of computational data, which burns their weekly quotas within minutes.
The shift in pricing model attracted backlash from developers, causing Google to expand usage limits for Gemini Pro models. The Gemini 3.5 Flash Low model is designed to fix Google’s Antigravity token limitation issue.
Multiple Variants of Gemini 3.5 Flash Model
Google appears to be developing different variants of the Gemini 3.5. Flash. The original Gemini 3.5 Flash, which Google launched during its annual I/O event last week, will now be referred to as Gemini 3.5 Flash (Medium).
The company is expected to release another variant, Gemini 3.5 Flash (High) that will handle complex tasks. Gemini 3.5 Flash is the first in the Gemini 3.5 family. Gemini 3.5 models are four times faster than frontier models and less costly.
Additionally, Google announced that it had updated Gemini token quotas across free and paid plans. The move is aimed at ensuring that users continue performing their software engineering tasks without disruptions.
User Concerns Around Rate Limit
Gemini users have also raised concerns about the rate limits on Google’s image generation model and asked the company to address that too.
“You need to match Codex or offer something better, I can generate 1000 images on Codex and can only generate 24 on Antigravity on the Ultra plan? Let me bring Flow inside Antigravity because Nano Banana V2 /Pro both are unlimited there,” One user posted on X.
In response, Mohan acknowledged that the limits were pretty low, adding that it “makes sense to increase” them. However, there was no commitment to boost the limits.
