Google Gemini 3.8 Flash has launched just weeks after its predecessor, with the company positioning the release as a more capable option for complex tasks. Google says the new model “works harder” than Gemini 3.7 Flash by performing additional reasoning steps and calling tools iteratively when handling demanding workloads.
The update arrives at a rapid cadence, following Gemini 3.7 Flash by only a short period. That pace reflects Google’s continued push to refine its lightweight Flash line, which is designed to balance speed with performance for developers building on the platform.
Pricing and Token Usage
Gemini 3.8 Flash carries the same introductory pricing as Gemini 3.7 Flash, set at roughly £0.60 per million input tokens and £2.95 per million output tokens. Despite the matching rates, Google warns that users could still end up paying more in practice. The company notes that the model might use more tokens to maximise performance, particularly at higher effort levels.
Because pricing is tied to token consumption, the additional reasoning steps and iterative tool calls that define the new model can increase the total number of tokens processed. That means a task completed on Gemini 3.8 Flash could cost more than the same task run on the earlier release, even though the per-token rates are identical.
Options for Developers
Developers who want to keep costs predictable are not required to migrate. Google confirms that Gemini 3.7 Flash remains available for those who prefer to minimise token usage, giving teams the choice between the newer model’s expanded reasoning and the older model’s tighter consumption profile.
The launch was followed by some early feedback as developers began testing the model in real workloads. The trade-off between capability and cost sits at the centre of the release, with the higher effort levels driving both the promised performance gains and the potential for increased token spend. For users prioritising raw output on complex tasks, Gemini 3.8 Flash offers more reasoning depth, while those focused on efficiency retain access to the previous version.
Gemini 3.8 Flash and Gemini 3.7 Flash share the same introductory rate of roughly £0.60 per million input tokens and £2.95 per million output tokens.
Source
Image: theverge.com
