Model Tracker

Gemini 3.1 Flash-Lite

Gemini 3.1 Flash-Lite is an AI model made by Google. Google released it on 3 March 2026. It is generally available.

In inglese

Gemini 3.1 Flash-Lite is an AI model made by Google. Google released it on 3 March 2026. It is generally available. It takes text, image, audio, video and returns text.

Released: 3 March 2026 — ai.google.dev, read 16 October 2026
Status: Generally available — ai.google.dev, read 16 October 2026
Takes: Text, Image, Audio, Video — deepmind.google, read 16 October 2026
Returns: Text — deepmind.google, read 16 October 2026
Open weights: Not confirmed: The model card lists distribution channels but does not state whether the model's weights are open.
Licence: Not confirmed: The cited model and API pages do not name a model licence.
Context window: Not confirmed: two readings disagree: 1048576 and 1000000
Longest answer: Not confirmed: two readings disagree: 65536 and 64000
Knowledge cutoff: Not confirmed: The cited maker pages do not print a knowledge-cutoff date for this model.
API price, input: Not confirmed: The standard pricing page gives different input prices by modality: $0.25 for text, image, and video, and $0.50 for audio. It does not print one standard input amount for the field.
API price, output: $1.50 per 1M tokens — ai.google.dev, read 16 October 2026
Where: Not confirmed: two readings disagree: api, app, cloud and api
Parameters: Not confirmed: The cited maker pages do not publish a parameter count for this model.
Model card: https://deepmind.google/models/model-cards/gemini-3-1-flash-lite/ — deepmind.google, read 16 October 2026

Benchmarks

Measured by others, not by the maker: each score with the name the measurer gives the model and the setting it ran, its day and its licence.

Epoch AI

Epoch Capabilities Index, Gemini 3.1 Flash-Lite: 144.5 (142.3–146.0), as of 19 October 2026
GPQA Diamond, gemini-3.1-flash-lite_high: 81.8% (± 2.7), run 6 August 2026
GPQA Diamond, gemini-3.1-flash-lite_low: 74.2% (± 3.1), run 6 August 2026
GPQA Diamond, gemini-3.1-flash-lite_minimal: 73.7% (± 3.1), run 6 August 2026
FrontierMath, tiers 1–3, gemini-3.1-flash-lite_high: 27.7% (± 2.7), run 30 August 2026
FrontierMath, tiers 1–3, gemini-3.1-flash-lite_low: 22.5% (± 2.5), run 30 August 2026
FrontierMath, tiers 1–3, gemini-3.1-flash-lite_minimal: 21.4% (± 2.4), run 28 August 2026
OTIS Mock AIME 2024–2025, gemini-3.1-flash-lite_high: 80.0% (± 6.0), run 6 August 2026
OTIS Mock AIME 2024–2025, gemini-3.1-flash-lite_low: 44.4% (± 7.5), run 6 August 2026
OTIS Mock AIME 2024–2025, gemini-3.1-flash-lite_minimal: 37.8% (± 7.3), run 6 August 2026

Epoch AI, ‘Capabilities & benchmarking’, epoch.ai/benchmarks, CC BY 4.0: Epoch AI’s own runs and its Capabilities Index. Data · CC BY 4.0

Questions

When was Gemini 3.1 Flash-Lite released?

Google released Gemini 3.1 Flash-Lite on 3 March 2026.

Is Gemini 3.1 Flash-Lite still available?

Yes: it is generally available, as Google states it.

What does Gemini 3.1 Flash-Lite take and return?

Gemini 3.1 Flash-Lite takes text, image, audio, video and returns text.

Sources

Gemini 3.1 Flash-Lite | Gemini API | Google AI for Developers (read 16 October 2026)
Release notes | Gemini API | Google AI for Developers (read 16 October 2026)
Gemini Developer API pricing | Gemini API | Google AI for Developers (read 16 October 2026)
Gemini 3.1 Flash-Lite - Model Card (read 16 October 2026)

All of the Model Tracker