GoogleGemini 2.5 Flash-Lite
Gemini 2.5 Flash-Lite is a Google proprietary model, released 17 Jun 2025. Its best configuration ranks #432 of 785 on the BenchLeader Index at 48.3 ±6.1, in the lower half. The ± is the point: 374 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in maths (58) and lowest in reasoning (39). At $0.175 per million tokens blended it is among the cheapest fifth of ranked models. It has been measured at 3 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 24 Nov 2025.stale: no new result in six months
- Blended price
- $0.175/M
- $0.100 in · $0.400 out
- Output speed
- –
- First answer
- –
- Context
- 1.0M
- Full answer
- –
- Cost per run
- –
- Released
- 17 Jun 2025
- Overall index48
- Reasoning39
- Coding40
- Agents & tools49
- Maths58
- Instruction following52
Versions
Google has shipped 6 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Gemini 3.5 Flash Lite | 21 Jul 2026 | 54.4 | #253 |
| Gemini 3.1 Flash Lite | 3 Mar 2026 | 53.7 | #274 |
| Gemini 2.5 Flash Lite 09thinking | 25 Sept 2025 | 47.9 | #443 |
| Gemini 2.5 Flash-Litethis page | 17 Jun 2025 | 48.3 | #432 |
| Gemini 2.0 Flash Lite 001 | 25 Feb 2025 | – | – |
| Gemini 2.0 Flash-Lite | 5 Feb 2025 | 47.2 | #462 |
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| no reasoning | 44.6 | #530 | 295 tok/s | 0.32 s | $0.0002 | Agents & tools 36 · Coding 52 · Composite 37 · Human preference 55 · Instruction following 35 · Knowledge 35 · Long context 40 · Maths 53 · Multimodal 44 · Reasoning 45 |
| thinking | 46.0 | #500 | 377 tok/s | 14 s | $0.0002 | Agents & tools 38 · Coding 45 · Composite 39 · Human preference 54 · Instruction following 51 · Knowledge 41 · Long context 39 · Maths 53 · Multimodal 45 · Reasoning 45 |
| not statedbest | 48.3 | #432 | – | – | $0.0002 | Agents & tools 49 · Coding 40 · Instruction following 52 · Maths 58 · Reasoning 39 |
How its index has moved
9 Sept 2026 to 21 Oct 2026- not stated
- thinking
- no reasoning
The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Arena Hard Prompts | 1391#179 | 1382#186 | – | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 47.4%#440 | 62.5%#360 | – | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 3.7%#577 | 6.8%#379 | – | Artificial Analysis | |
| Kagi LLM Benchmarkeffort not stated | – | – | 40.5%#103 | Kagi LLM Benchmark | |
| GPQA Diamond (HELM)not in indexeffort not stated | – | – | 30.9%#60 | HELM Capabilities | |
| CritPt | 0.0%#315 | 0.0%#315 | – | Artificial Analysis | |
| LMCAnot in index | – | 18.1%#163 | – | Epoch AI Benchmarking Hub | |
| DTBenchnot in index | – | 62.8%#173 | – | Epoch AI Benchmarking Hub |
Coding
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| WeirdML | – | 35.2%#125 | 35.2%#125 | WeirdML | |
| Arena Coding | 1398#197 | 1382#214 | – | Arena (LMArena) | |
| ALE-Benchnot in index | – | 325.9#115 | – | Epoch AI Benchmarking Hub |
Agents & tools
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Terminal-Bench Hard | 2.3%#320 | 4.5%#284 | – | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 19.0%#343 | 18.4%#347 | – | Artificial Analysis | |
| BFCL Overalleffort not stated | – | – | 36.9%#40 | Berkeley Function Calling Leaderboard |
Maths
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Omni-MATH (HELM)effort not stated | – | – | 48.0%#20 | HELM Capabilities | |
| Arena Maths | 1362#193 | 1363#185 | – | Arena (LMArena) |
Knowledge
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| AA-Omniscience | -58.8#456 | -45.6#354 | – | Artificial Analysis | |
| MMLU-Pro (HELM)not in indexeffort not stated | – | – | 53.7%#55 | HELM Capabilities | |
| MedCodenot in indexeffort not stated | – | – | 27.1%#98 | Vals AI | |
| MedScribenot in indexeffort not stated | – | – | 72.8%#83 | Vals AI | |
| AA-Omniscience: accuracynot in index | 15.4%#413 | 17.9%#351 | – | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 12.2%#377 | 22.6%#263 | – | Artificial Analysis |
Instruction following
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| IFBench | 31.5%#351 | 49.9%#175 | – | Artificial Analysis | |
| IFEval (HELM)effort not stated | – | – | 81.0%#44 | HELM Capabilities | |
| WildBench (HELM)effort not stated | – | – | 81.8%#24 | HELM Capabilities | |
| Arena Instruction Followingnot in index | 1364#179 | 1369#173 | – | Arena (LMArena) |
Human preference
| Benchmark | no reasoning | thinking | Source | Trend |
|---|---|---|---|---|
| Arena Text | 1380#174 | 1375#178 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1360#159 | 1372#145 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1373#176 | 1366#183 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1387#172 | 1379#179 | Arena (LMArena) |
Multimodal
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Arena Vision | 1198#92 | 1188#98 | – | Arena (LMArena) | |
| MMMU-Pro | 54.0%#231 | 58.2%#219 | – | Artificial Analysis | |
| SAGEnot in indexeffort not stated | – | – | 27.5%#84 | Vals AI |
Long context
| Benchmark | no reasoning | thinking | Source | Trend |
|---|---|---|---|---|
| Fiction.LiveBench 120k | – | 21.9%#33 | Fiction.live | |
| AA-LCR | 32.0%#387 | 55.7%#296 | Artificial Analysis |
Composite
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Epoch Capabilities Indexnot in indexeffort not stated | – | – | 133.8#137 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index v4.3.2 | 6.7#513 | 8.6#434 | – | Artificial Analysis | |
| HELM Capabilities meannot in indexeffort not stated | – | – | 59.1%#46 | HELM Capabilities |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.010 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0002 | $0.0001 | – |
| Summarize a 30-page report | 12,000 / 600 | $0.0014 | $0.0006 | – |
| Code edit | 6,000 / 1,500 | $0.0012 | $0.0008 | – |
| Agentic coding session | 60,000 / 4,000 | $0.0076 | $0.0036 | – |
| Structured extraction | 2,000 / 200 | $0.0003 | $0.0001 | – |
More from Google
See also
Questions people ask
- What are Gemini 2.5 Flash-Lite's benchmark scores?
- As of 25 Oct 2026, Gemini 2.5 Flash-Lite scores 48.3 ±6.1 on the BenchLeader Index, strongest in maths (58) and weakest in reasoning (39). That rests on 6 independent benchmark results, and every score on this page links to its publisher.
- How does Gemini 2.5 Flash-Lite rank?
- Gemini 2.5 Flash-Lite is #432 of 785 ranked configurations on 25 Oct 2026. 374 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does Gemini 2.5 Flash-Lite cost per 1M tokens?
- As of 25 Oct 2026, Gemini 2.5 Flash-Lite costs $0.100 per 1M input tokens and $0.400 per 1M output tokens, $0.175 blended (three input tokens per output token).
- When was Gemini 2.5 Flash-Lite released?
- Google released Gemini 2.5 Flash-Lite on 17 Jun 2025. BenchLeader has results for it at 3 reasoning-effort settings.
Data as of 25 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Gemini 2.5 Flash-Lite: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/gemini-2-5-flash-lite.