Gemini Flash Lite
Google's lightweight, low-cost Gemini model tier for high-volume, latency-sensitive tasks.
HN commenters like it as a cheap, fast model12 for classification, summarization and other production workloads.34567 The most common complaint is Google discontinuing older versions such as 2.5 Flash Lite,8910 which forces migrations and, for some, higher prices.1112
Why people recommend it
Production workloads · 3
Used in production for classification, image processing, summarization and as a primary German-language model.
“in production we have been using Gemini Flash Lite as primary”
“3.5 Flash Lite sounds like it could be a strict upgrade for our use case”
Show 1 more
“I had a production image processing using Gemini 2.5 Flash Lite”
Good value · 2
Described as surprisingly good value with a strong speed, accuracy and price tradeoff.
“Gemini Flash Lite 3.1 is however surprisingly good value.”
“3.5 Flash Lite sounds like it could be a strict upgrade for our use case”
Capable enough · 2
Commenters say it handles a lot and is all they need for small apps.
“You can do a lot of things with Gemini Flash Lite 3.1.”
“please please don't kill the old ones like gemini-2.5-flash-lite, because that all I ever need for my little apps”
Classification · 2
Called fast, cheap and accurate for classifying eCommerce listings.
“3.5 Flash Lite sounds like it could be a strict upgrade for our use case”
“classification tasks that I used Gemini 2.5 flash lite with”
Caveats and dissent
Model deprecation · 3
Older versions like 2.5 Flash Lite are being discontinued, forcing users to migrate or switch providers.
“please please don't kill the old ones like gemini-2.5-flash-lite, because that all I ever need for my little apps”
“We spent thousands just to build this workflow now Google says screw off”
Show 1 more
“I had a production image processing using Gemini 2.5 Flash Lite”
Price increases · 1
One developer says pricing increases forced them to raise prices and cut limits.
“please please don't kill the old ones like gemini-2.5-flash-lite, because that all I ever need for my little apps”
Free quota removed · 1
A commenter shut down a public instance after the generous free quota was removed.
“that's what I used for by LLM-backed HTTP server”
Slow releases · 1
One commenter noted the Flash Lite line lagged behind schedule at version 2.5.
“I use Gemini flash lite in a side project, and it’s stuck on 2.5.”
Often compared with
Who recommends it in their own words
Threads where it came up
- Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber 3 mentions
- Different Gemini model setups for agentic video analysis: Deep vs. Lite 1 mention
- Gemini 4 Argon 1 mention
- I built non-autoregressive decision models with RL a year ago 1 mention
- Qwen 3.8 Omni Flash 1 mention
- Qwen 3.8 27B is out: open weights, best local dense model yet 1 mention
Also written as Gemini 2.5 Flash Lite, Gemini Flash Lite 3.1, Gemini 3.5 Flash Lite, Gemini 3.5 Flash-Lite. Data: gemini-flash-lite.json. Quotes are verbatim; the verdict and theme labels are written by a model from these comments, and every clause links to its sources.
All comments 15
“a full analysis of Gemini 3.8 Flash vs Gemini 3.5 Flash lite using agentic video understanding”
“mine has 3.1 Pro, 3.5 Flash-Lite, and 3.8 Flash (with optional extra effort)”
“classification tasks that I used Gemini 2.5 flash lite with”
“I had a production image processing using Gemini 2.5 Flash Lite”
“in production we have been using Gemini Flash Lite as primary”
“I use Gemini Flash Lite to summarize and filter the bulletins.”
“please please don't kill the old ones like gemini-2.5-flash-lite, because that all I ever need for my little apps”
“We spent thousands just to build this workflow now Google says screw off”
“3.5 Flash Lite sounds like it could be a strict upgrade for our use case”
“Gemini Flash Lite 3.1 is however surprisingly good value.”
“You can do a lot of things with Gemini Flash Lite 3.1.”
“It's mostly tested on small analysis and question answering with Gemini Flash Lite”
“It made me even use the tiny models like Gemini Flash Lite and GPT Mini/Nano more too.”
“that's what I used for by LLM-backed HTTP server”
“I use Gemini flash lite in a side project, and it’s stuck on 2.5.”