Gemini 3.1 Flash Lite

May 2026
Retires May 7, 2027Move toGemini 3.5 Flash LiteNewer version
Gemini 3.1 Flash Lite
Compare in
on
$0.06per task?
$0.25 / $1.50per 1M in / outcache $0.025
362 tok/sfirst token in 4.7 s?
1Mcontext

Rankingat high effort?

Strong at

S
Instructions#11 of 245
B
Coding#65 of 134Reasoning#136 of 336Science#103 of 312Long context#95 of 314Knowledge#113 of 320Vision#43 of 125
C
Agents & work tasks#107 of 177Tool calling#147 of 242

Alternatives?

ModelScoreCostPer taskWhat changes
CheaperAgnes 2.5 Pro Beta+10%-73%$0.10 / $0.30BetterGPT-5.6 Lunamedium+4%-74%$0.016OpenKimi K3max+14%+3,224%$2.00Sampling parameters are fixed+3CheaperHy3 previewthinkingOpen+5%-49%$0.18 / $0.60Context: 1M → 256K+3CheaperLing 3.0 Flash VLthinkingOpen+5%-84%$0.06 / $0.18Context: 1M → 262K+1

History

4 versions
606570758085Oct 25Jan 26Apr 26Jul 26Release dateGPQA Diamond2.53.13.5
606570758085Jan 26Jul 26Release dateGPQA Diamond2.53.13.5

Benchmarks

all categories
Chat & Knowledge
Humanity's Last Exam17.2#136
AA Omniscience Index-16.4#114
CritPt (physics)1.1#122
AA Intelligence Index 4.3.215.6#144
AA Omniscience · Accuracy36.3#58
AA Omniscience · Hallucination Rate82.7#188
AA-LCR (long context) 1.174.3#95
GDP.pdf · AA all-pass7.8#72
MLCR-AA5#79
FrontierMath Tiers 1–3 (v2)27.7#59
GPQA Diamond82.2#104
GPQA Diamond · Epoch81.8#67
OTIS Mock AIME 2024 202580#72
Coding
Terminal-Bench 4.00.5#65
SciCode43.4#82
Terminal-Bench Hard (version unstated)24.2#77
Terminal-Bench 2.131.1#95
Agents & Tool Use
AutomationBench-AA6.8#80
GDPval-AA (work tasks) 2.1430#107
MCP-Atlas · full 100057.1#12
MCP-Atlas Public60.4#13
τ³-Banking · AA9.7#92
AA Analyst Agent8.8#40
APEX-Agents12.2#23
EnterpriseOps-Gym-AA28#39
AA-Briefcase 1.1 · Elo207.1#119
τ²-bench · AA31.3#149
Writing & Language
IFBench (instructions)77.2#11
Vision & Documents
MMMU-Pro (vision)75.5#43

FAQ

How much does Gemini 3.1 Flash Lite cost?›

$0.25 per 1M input tokens and $1.50 per 1M output tokens; cached input $0.025 per 1M. A typical benchmark task costs about $0.06 at high effort.

Is Gemini 3.1 Flash Lite free to use?›

There is no free API; the cheapest paid price is $0.25 per 1M input tokens.

How good is Gemini 3.1 Flash Lite, and what is it best at?›

#136 of 336 in Chat & Knowledge, #65 of 134 in Coding, #80 of 135 in Agents & Tool Use, #11 of 245 in Writing & Language, #43 of 125 in Vision & Documents. Strongest at: Instructions.

Is Gemini 3.1 Flash Lite better than GPT-5.5 Instant (June 2026)?›

On GPQA Diamond, Gemini 3.1 Flash Lite scores 82.2 vs 82.3 — about the same. It is 89% cheaper per task.

How can I use Gemini 3.1 Flash Lite?›

Through the API of DeepInfra, Google, Google AI Studio.

What is the context window of Gemini 3.1 Flash Lite?›

1M tokens, with up to 66K output tokens.

When was Gemini 3.1 Flash Lite released? Is there a newer version?›

Released 2026-05-07. It has been superseded by Gemini 3.5 Flash Lite.

Is Gemini 3.1 Flash Lite open source?›

No, Gemini 3.1 Flash Lite is a closed model available through an API. The closest open-weights option here is Kimi K3.

Can Gemini 3.1 Flash Lite read images?›

It accepts text and images, video as input.

What is Gemini 3.1 Flash Lite?›

Gemini 3.1 Flash Lite is a closed language model from Google, released May 2026, the successor to Gemini 3.1 Flash Lite Preview.

What are the best alternatives to Gemini 3.1 Flash Lite?›

Agnes 2.5 Pro Beta (cheaper), GPT-5.6 Luna (better), Kimi K3 (open weights).