AI Index

Gemini 3.5 Flash-Lite

Language Google Verified Released Jul 21, 2026
  • In TextImageVideoAudioPDF Out Text
  • Reasoning· minimal, low, medium, high
  • Tool use
  • Structured output

Prices

3 offerings · prices as listed
Gemini API Official Global · USD gemini-3.5-flash-lite
Conditions/ 1M tokensInputCache readOutput
Standard —$0.030$2.50
Text tokens $0.30——
Image tokens $0.30——
Audio tokens $0.30——
Video tokens $0.30——
Flex tier —$0.020$1.25
Priority tier —$0.050$4.50
Batch —$0.015$1.25
Text tokensFlex tier $0.15——
Text tokensPriority tier $0.54——
Text tokensBatch $0.15——
Image tokensFlex tier $0.15——
Image tokensPriority tier $0.54——
Image tokensBatch $0.15——
Audio tokensFlex tier $0.15——
Audio tokensPriority tier $0.54——
Audio tokensBatch $0.15——
Video tokensFlex tier $0.15——
Video tokensPriority tier $0.54——
Video tokensBatch $0.15——

Prices from Gemini API pricing (checked Nov 10, 2026)

Standard: the price a request with no conditions pays (the headline price elsewhere on the site).

Details

Max input
1M
Built-in tools
Web search, Web fetch, Code execution, Computer use, Maps
Knowledge cutoff
Mar 2026
Open weights
—
Family
gemini-flash-lite
Model ID

Not stated on Google’s pages: reasoning can be off, max thinking budget, aliases.

Cited scores

From third-party leaderboards, not measured by us
  • Epoch Capabilities Index 145.6

    #77 of 118 listed models · CI 143.9–147.2

  • LMArena Text 1,455

    #49 of 163 listed models · CI 1,450–1,459 · 35,133 votes

  • Epoch AI · Chess Puzzles 21%

    #62 of 85 listed models · low effort · ±4.09

    Also listed: high effort 16% · minimal effort 7%

  • Epoch AI · FrontierMath Tier 4 v2 0%

    #63 of 65 listed models · high effort · ±0

  • Epoch AI, 'Capabilities & benchmarking'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/benchmarks'. Licensed CC BY 4.0. · License · as of Oct 28 – Nov 5, 2026
  • LMArena (Arena), leaderboard-dataset on Hugging Face, licensed CC BY 4.0 · License · as of Nov 8, 2026
  • EQ-Bench by Samuel J. Paech (eqbench.com; EQ-bench/EQ-bench-site on GitHub), MIT License. · License · as of Oct 28, 2026
  • LiveBench (White et al.), release 2026-06-25, livebench.ai, licensed CC BY-SA 4.0. This board, adapted from it, is shared under CC BY-SA 4.0. · License · as of Jun 25, 2026
  • changes we made

Provider-reported scores

Published by Google, under the conditions it states
BenchmarkSettingScore
SWE-bench Propublic set, internal Antigravity harness, thinking high54.2%
Terminal-Bench 2.1Terminus-2 harness, thinking high54%
OSWorld-Verifiedaverage of 5 runs, single attempt, max 100 steps, thinking high74%
CharXiv Reasoningno tools, thinking high74.5%
with tools (search + code execution), thinking high76.5%

News

All news →
  • Sep 1, 2026Capability Google adds agentic video understanding to Gemini Agentic video understanding is available for Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite on the Interactions and GenerateContent APIs. Google says it can use up to 88% fewer tokens on long-form video than static processing.

Changes

All changes →
  1. Cache read (Batch) price cut 25% $0.020 → $0.015 / 1M tokens · Gemini API

Sources

Google’s official policies: data use, privacy, terms →

Google · United States · ai.google.dev

Snapshots, the same family and other models from this provider