Skip to content
OSMEXO

Gemini 3.5 Flash Lite

google/gemini-3.5-flash-lite
  • Google
  • Family · Gemini 3 Flash Lite
  • Released 21 Jul 2026
  • Text · Image · Video · File · Audio → Text
  • Reasoning · adjustable
  • Tools · JSON · Vision
Spec & list price
Context1M
Max output64K
Input / 1M$0.30
Output / 1M$2.508.3× input
Scores · standing
Intelligence22.2#120/489
Coding49.3#69/196
Usage
Tokens / day25.2B▲ 2.1% 7d
Usage rank#66
Requests / day4.3M
Tool calls / req0.02
Per request · 30 days
Avg input2.1K0.1× median
Avg output1020.1× median
Cache hit30%0.4× median
Reasoning share14%0.2× median
Tool-call failures3.1%1.6× median

Benchmarks

bar = share of the top score · Benchmark data: Artificial Analysis
GPQA Diamond#107 of 445
83.8%
Long-context reasoning#93 of 393
76.0%
SciCode#102 of 160
41.3%
Coding index#69 of 196
49.3
Terminal-Bench 2.1#71 of 182
53.6%
Intelligence index#120 of 489
22.2
τ-Bench Banking#67 of 154
17.5%
Humanity's Last Exam#149 of 465
18.8%
Speed · median across 8 providers, latest day
Output speed#37/333130 tok/s1.9× faster than median
Fastest provider209 tok/sbest of 8 providers
First token#1470.9 s1.1× quicker than median
First answer token#344.66 sthinking included

#17 cheapest of 31 with a similar score

Value
Intelligence Index
15202530$0.2$0.5$1Gemini 3.5 Flash Lite
Blended price per 1M tokens (log)
Family progress
Intelligence Index
141618202224AprMayJunJulGemini 3.5 Flash Lite
  • Family releases22.2

Technical details

Reasoning
on · minimal
Tokenizer
Gemini
Moderation
none
Reasoning effort
  • high
  • medium
  • low
  • minimal · default
Request parameters 12
  • tools
  • tool_choice
  • response_format
  • structured_outputs
  • reasoning
  • reasoning_effort
  • include_reasoning
  • temperature
  • top_p
  • max_tokens
+2 more
  • stop
  • seed

Interesting facts

  1. Less output room than similar models (65.5K against a 129.5K median).
  2. Released 103 days ago, the newest of 2 Gemini 3 Flash Lite models.
  3. 2nd of 3 Gemini 3 Flash Lite models by tokens last month.

When to use

Best for

  • Low-cost bulk classification and extraction
  • High-volume chat where latency matters
  • Tagging images, audio and video at scale
  • Long document summarization on a budget

Think twice for

  • Hardest reasoning and coding tasks
  • Complex agent workflows
  • Self-hosting (closed weights)

Research papers

Same family

Gemini 3 Flash Lite

Alternatives

other vendors · similar capability

Using Gemini 3.5 Flash Lite at volume?

Sign in to ask for a volume discount. You will continue in the OSMEXO dashboard.