Gemini 3.5 Flash-Lite
No shutdown before this date. Gemini 3.5 Flash-Lite stays on Google Vertex AI until at least Jul 21, 2027.
≥ 9 mo- Earliest retirement
- ≥ 280 daysJul 21, 2027 · Google Vertex AI
- Arena Text
- 1,455#38 of 128 · Arena
- GPQA Diamond
- 83.3%#56 of 108 · Epoch AI
- List price · $ per 1M
- $0.30 in $2.50 outGemini API · cached input $0.03
- Context window
- 1MMax output 65.5K tokens
Dates on every platform
| Platform | Model ID and status | Deprecated | Shuts down | Replacement |
|---|---|---|---|---|
| Gemini APIGoogle: Gemini deprecations (opens in a new tab) · checked Oct 14, 2026 | gemini-3.5-flash-liteActive | – | None set | – |
| Google Vertex AIGoogle Cloud: Model versions and lifecycle (opens in a new tab) · checked Oct 14, 2026 | gemini-3.5-flash-liteActive | – | ≥ 9 moNot sooner than | – |
What breaks when you move off Gemini 3.5 Flash-Lite
- Parameters
- Parameters
- Parameters
- Parameters
- Parameters
- Defaults
Intelligence
| Benchmark | Publisher | Score | Rank | Published |
|---|---|---|---|---|
| Arena Text | Arena | 1,455 ± 2 | #38 of 128 | Oct 8, 2026 |
| GPQA Diamond | Epoch AI | 83.3% ± 2.7% | #56 of 108 | Aug 6, 2026 |
| Arena WebDev | Arena | 1,440 ± 21 | #39 of 69 | Oct 8, 2026 |
| FrontierMath Tier 4 | Epoch AI | 0.0% ± 0.0% | #57 of 58 | Aug 2, 2026 |
| FrontierMath Tiers 1-3 | Epoch AI | 26.0% ± 2.6% | #59 of 71 | Aug 2, 2026 |
| OTIS Mock AIME 2024-2025 | Epoch AI | 71.1% ± 6.8% | #69 of 105 | Aug 6, 2026 |
Arena (arena.ai), https://arena.ai/leaderboard
Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource].
List prices
| Platform | Input | Cached | Output | In effect | Source |
|---|---|---|---|---|---|
Gemini APIgemini-3.5-flash-lite | $0.30 | $0.03 | $2.50 | Jul 21, 2026Current | Google: Gemini Developer API pricing (opens in a new tab) |
Google Vertex AIgemini-3.5-flash-lite | $0.30 | $0.03 | $2.50 | Jul 21, 2026Current · Global endpoint price; regional endpoints cost more. | Google Cloud: Agent Platform pricing (opens in a new tab) |
Change history
ShutdownGemini 3.1 Flash-Lite can shut down from May 7, 2027 on the Gemini APIGemini API
Google lists May 7, 2027 as the earliest shutdown date for gemini-3.1-flash-lite and recommends gemini-3.5-flash-lite.
- Platform
- Gemini API
- Replacement
- Gemini 3.5 Flash-Lite
API changeGemini deprecates temperature, top_p and top_kGemini API2 breaks
From Gemini 3.6 Flash and Gemini 3.5 Flash-Lite on, the Gemini API ignores sampling parameters and rejects prefilled model turns. The change applies to later Gemini models too.
- Platform
- Gemini API
- Parameterstemperature, top_p and top_k are ignored. Google says later model generations will return an error.
- ParametersA request whose last non-empty turn is a model turn returns a 400 error.
Source: Google, Gemini API release notes (opens in a new tab); Google, What's new in Gemini 3.6 Flash and 3.5 Flash-Lite (opens in a new tab)
ReleaseGemini 3.6 Flash and Gemini 3.5 Flash-Lite releasedGemini API, Google Vertex AI4 breaks
Google released Gemini 3.6 Flash and Gemini 3.5 Flash-Lite as generally available, with new rules for sampling parameters and model turns.
- Platform
- Gemini API, Google Vertex AI
- ParametersCustom temperature, top_p and top_k values are ignored.
- ParametersOn Agent Platform, custom frequency and presence penalty values return an API error.
- ParametersA request whose last turn is a model turn returns an error.
- DefaultsGemini 3.5 Flash-Lite defaults to the minimal thinking level; set medium or high for agent work.
Source: Google, Gemini API release notes (opens in a new tab); Google, What's new in Gemini 3.6 Flash and 3.5 Flash-Lite (opens in a new tab); Google Cloud, Gemini Enterprise Agent Platform release notes (opens in a new tab)