FRENES

Google · language model

Gemini 3.1 Flash Lite Preview: pricing, context and benchmarks

Gemini 3.1 Flash Lite Preview is a language model from Google released on March 3, 2026. It costs $0.25 per million input tokens and $1.5 per million output tokens, with a 1M-token context window. Epoch AI ranks it #87 of 125 rated models.

Score
55/100
Performance (ECI) 13/40
Price 17/30
Context 14/15
Features 11/15
Mid-range costReasoningTool calling

What you need to know about Gemini 3.1 Flash Lite Preview

Gemini 3.1 Flash Lite Preview was released by Google (United States) on March 3, 2026. It accepts text, images, video, files, and audio as input and produces text. Its context window reaches 1,048,576 tokens, larger than 93 % of the models in the registry, and it can write up to 65,536 tokens in one go.

It has been available for about 8 months, long enough for its prices and scores to have settled.

Gemini 3.1 Flash Lite Preview pricing

On OpenRouter, Gemini 3.1 Flash Lite Preview costs $0.25 per million input tokens and $1.5 per million output tokens. Counting three tokens read for every token written, its average price is $0.563 per million, more expensive than 44 % of paid models (1.4 times cheaper than the median of $0.76).

A concrete example: a task that sends 2,000 tokens (about 1,500 words) and receives a 500-token answer costs $0.0013. At 1,000 tasks a day, budget $38 a month. The price calculator redoes this with your own volumes.

Tokens already sent and cached are billed at $0.025 per million.

Performance and benchmarks

The Epoch Capabilities Index (ECI), which aggregates published results across many benchmarks, gives Gemini 3.1 Flash Lite Preview a score of 144.3 (range 142.3 to 145.7). That places it #87 of 125 rated models.

Scores published by Epoch AI on the benchmarks the site tracks:

Source: Epoch AI, Benchmarking Hub (CC BY 4.0).
BenchmarkWhat it measuresScoreDate
GPQA DiamondPhD-level science questions81.8 %Mar 3, 2026
FrontierMathresearch-level maths problems (tiers 1-3)27.7 %Mar 3, 2026
OTIS Mock AIMEmaths competition problems80 %Mar 3, 2026
GPQA DiamondGPQA Diamond : 81.8 %81.8 %FrontierMathFrontierMath : 27.7 %27.7 %OTIS Mock AIMEOTIS Mock AIME : 80 %80 %
Gemini 3.1 Flash Lite Preview success rate by benchmark (Epoch AI).
116129142156169$0.1$1$10Average price ($ per million tokens, log scale)ECIClaude Opus 5.5 : Average price ($ per million tokens, log scale) $8, ECI 169GPT-6 Astra : Average price ($ per million tokens, log scale) $20, ECI 169GPT-6.1 Sol : Average price ($ per million tokens, log scale) $4, ECI 168Claude Fable 5.1 : Average price ($ per million tokens, log scale) $20, ECI 166Gemini 4 Argon : Average price ($ per million tokens, log scale) $4, ECI 166Claude Sonnet 5.5 : Average price ($ per million tokens, log scale) $4, ECI 166Claude Opus 5 : Average price ($ per million tokens, log scale) $10, ECI 164GPT-6 Sol : Average price ($ per million tokens, log scale) $4, ECI 164Claude Fable 5 : Average price ($ per million tokens, log scale) $20, ECI 163GPT-5.5 Pro : Average price ($ per million tokens, log scale) $68, ECI 163GPT-5.6 Sol : Average price ($ per million tokens, log scale) $8, ECI 163Muse Wonder : Average price ($ per million tokens, log scale) $8, ECI 162GPT-5.6 Terra : Average price ($ per million tokens, log scale) $5, ECI 160GPT-5.5 : Average price ($ per million tokens, log scale) $11, ECI 160GPT-5.4 Pro : Average price ($ per million tokens, log scale) $68, ECI 160Claude Haiku 5.5 : Average price ($ per million tokens, log scale) $0.2, ECI 159Claude Opus 4.8 : Average price ($ per million tokens, log scale) $10, ECI 159Gemini 3.7 Flash : Average price ($ per million tokens, log scale) $2, ECI 158Kimi K3 : Average price ($ per million tokens, log scale) $3, ECI 158Gemini 3.8 Flash : Average price ($ per million tokens, log scale) $2, ECI 157GPT-5.4 : Average price ($ per million tokens, log scale) $6, ECI 157Muse Spark 1.3 : Average price ($ per million tokens, log scale) $2, ECI 157GPT-5.3-Codex : Average price ($ per million tokens, log scale) $5, ECI 157Claude Sonnet 5 : Average price ($ per million tokens, log scale) $4, ECI 157Claude Opus 4.7 : Average price ($ per million tokens, log scale) $10, ECI 157GPT-6 Luna : Average price ($ per million tokens, log scale) $0.2, ECI 157Grok 4.6 : Average price ($ per million tokens, log scale) $3, ECI 157GPT-5.6 Luna : Average price ($ per million tokens, log scale) $0.5, ECI 157GPT-5.2 Pro : Average price ($ per million tokens, log scale) $58, ECI 156GLM 5.3 : Average price ($ per million tokens, log scale) $0.4, ECI 156DeepSeek V4.1 Flash 1017 : Average price ($ per million tokens, log scale) $0.3, ECI 156DeepSeek V4.1 Flash : Average price ($ per million tokens, log scale) $0.3, ECI 156Claude Opus 4.6 : Average price ($ per million tokens, log scale) $10, ECI 156DeepSeek V4 Pro 0813 : Average price ($ per million tokens, log scale) $1, ECI 156DeepSeek V4 Pro 0423 : Average price ($ per million tokens, log scale) $0.3, ECI 156Gemini 3.1 Pro Preview : Average price ($ per million tokens, log scale) $5, ECI 155Muse Spark Open : Average price ($ per million tokens, log scale) $2, ECI 155Qwen3.8 Max (0902) : Average price ($ per million tokens, log scale) $3, ECI 155Muse Spark 1.2 : Average price ($ per million tokens, log scale) $2, ECI 155Gemini 3.5 Flash : Average price ($ per million tokens, log scale) $3, ECI 155DeepSeek V4 Flash 0731 : Average price ($ per million tokens, log scale) $0.07, ECI 155DeepSeek V4 Flash 0423 : Average price ($ per million tokens, log scale) $0.1, ECI 155Gemini 3.6 Flash : Average price ($ per million tokens, log scale) $2, ECI 155Grok 4.7 : Average price ($ per million tokens, log scale) $3, ECI 154Grok 4.5 : Average price ($ per million tokens, log scale) $3, ECI 154Muse Spark 1.1 : Average price ($ per million tokens, log scale) $2, ECI 154Qwen3.7 Max : Average price ($ per million tokens, log scale) $2, ECI 154GPT-5.2 : Average price ($ per million tokens, log scale) $5, ECI 154GLM 5.3 Flash : Average price ($ per million tokens, log scale) $0.2, ECI 153Claude Sonnet 4.6 : Average price ($ per million tokens, log scale) $6, ECI 153Grok 4.20 : Average price ($ per million tokens, log scale) $2, ECI 152GLM 5.2 : Average price ($ per million tokens, log scale) $2, ECI 152Gemini 3 Flash Preview : Average price ($ per million tokens, log scale) $1, ECI 152Qwen3.8 27B : Average price ($ per million tokens, log scale) $0.8, ECI 152Kimi K2.6 : Average price ($ per million tokens, log scale) $1, ECI 151GPT-5 Pro : Average price ($ per million tokens, log scale) $41, ECI 151Inkling Small : Average price ($ per million tokens, log scale) $0.6, ECI 150Kimi K2.7 Code : Average price ($ per million tokens, log scale) $1, ECI 150Claude Opus 4.5 : Average price ($ per million tokens, log scale) $10, ECI 150Mistral Large 4 : Average price ($ per million tokens, log scale) $2, ECI 150GPT-5 : Average price ($ per million tokens, log scale) $3, ECI 150GLM 5.1 : Average price ($ per million tokens, log scale) $2, ECI 150GPT-5.1 : Average price ($ per million tokens, log scale) $3, ECI 150GPT-5.4 Mini : Average price ($ per million tokens, log scale) $2, ECI 149Inkling : Average price ($ per million tokens, log scale) $2, ECI 148Kimi K2.5 : Average price ($ per million tokens, log scale) $1, ECI 148o3 Pro : Average price ($ per million tokens, log scale) $35, ECI 147Qwen3.6 Plus : Average price ($ per million tokens, log scale) $0.7, ECI 147Qwen3.7 Plus : Average price ($ per million tokens, log scale) $0.6, ECI 147MiniMax M3 : Average price ($ per million tokens, log scale) $0.5, ECI 147Qwen3.5 397B A17B : Average price ($ per million tokens, log scale) $1, ECI 147Claude Sonnet 4.5 : Average price ($ per million tokens, log scale) $6, ECI 147o3 : Average price ($ per million tokens, log scale) $4, ECI 147MiniMax M2.5 : Average price ($ per million tokens, log scale) $0.5, ECI 147Qwen3.6 27B : Average price ($ per million tokens, log scale) $0.7, ECI 146DeepSeek V3.2 : Average price ($ per million tokens, log scale) $0.3, ECI 146GPT-5.4 Nano : Average price ($ per million tokens, log scale) $0.5, ECI 146Nemotron 3 Ultra : Average price ($ per million tokens, log scale) $1, ECI 146Kimi K2 Thinking : Average price ($ per million tokens, log scale) $1, ECI 146GLM 5 : Average price ($ per million tokens, log scale) $2, ECI 146MiniMax M2.7 : Average price ($ per million tokens, log scale) $0.5, ECI 146o4 Mini : Average price ($ per million tokens, log scale) $2, ECI 145GPT-5 Mini : Average price ($ per million tokens, log scale) $0.7, ECI 145Gemini 3.5 Flash Lite : Average price ($ per million tokens, log scale) $0.8, ECI 145DeepSeek V3.2 Exp : Average price ($ per million tokens, log scale) $0.3, ECI 145Gemini 3.1 Flash Lite : Average price ($ per million tokens, log scale) $0.6, ECI 144Claude Opus 4.1 : Average price ($ per million tokens, log scale) $30, ECI 144Qwen3.7 Flash : Average price ($ per million tokens, log scale) $0.06, ECI 144Qwen3.6 35B A3B : Average price ($ per million tokens, log scale) $0.3, ECI 144GLM 4.7 : Average price ($ per million tokens, log scale) $0.9, ECI 144Gemma 4 31B : Average price ($ per million tokens, log scale) $0.3, ECI 143Qwen3.6 Flash : Average price ($ per million tokens, log scale) $0.4, ECI 143Qwen3.5-35B-A3B : Average price ($ per million tokens, log scale) $0.4, ECI 142Claude Haiku 4.5 : Average price ($ per million tokens, log scale) $2, ECI 142Gemma 4 26B A4B : Average price ($ per million tokens, log scale) $0.1, ECI 142o1 : Average price ($ per million tokens, log scale) $26, ECI 142Mistral Medium 3.5 : Average price ($ per million tokens, log scale) $3, ECI 141o3 Mini : Average price ($ per million tokens, log scale) $2, ECI 140DeepSeek V3.1 : Average price ($ per million tokens, log scale) $0.4, ECI 140gpt-oss-120b : Average price ($ per million tokens, log scale) $0.3, ECI 140Qwen3.5-9B : Average price ($ per million tokens, log scale) $0.2, ECI 139GPT-5 Nano : Average price ($ per million tokens, log scale) $0.1, ECI 139R1 0528 : Average price ($ per million tokens, log scale) $0.9, ECI 139Qwen3 32B : Average price ($ per million tokens, log scale) $0.1, ECI 138Qwen3 14B : Average price ($ per million tokens, log scale) $0.1, ECI 138gpt-oss-20b : Average price ($ per million tokens, log scale) $0.04, ECI 138GPT-4.1 : Average price ($ per million tokens, log scale) $4, ECI 137Qwen3 30B A3B : Average price ($ per million tokens, log scale) $0.2, ECI 136GPT-4.1 Mini : Average price ($ per million tokens, log scale) $0.7, ECI 135Mistral Medium 3 : Average price ($ per million tokens, log scale) $0.8, ECI 134DeepSeek V3 0324 : Average price ($ per million tokens, log scale) $0.5, ECI 132DeepSeek V3 : Average price ($ per million tokens, log scale) $0.5, ECI 132Llama 4 Maverick : Average price ($ per million tokens, log scale) $0.3, ECI 132Phi 4 : Average price ($ per million tokens, log scale) $0.09, ECI 130Gemma 3 27B : Average price ($ per million tokens, log scale) $0.1, ECI 130Llama 4 Scout : Average price ($ per million tokens, log scale) $0.1, ECI 130GPT-4.1 Nano : Average price ($ per million tokens, log scale) $0.2, ECI 130GPT-4o-mini : Average price ($ per million tokens, log scale) $0.3, ECI 126GPT-4o-mini (2024-07-18) : Average price ($ per million tokens, log scale) $0.3, ECI 126Gemma 3 12B : Average price ($ per million tokens, log scale) $0.07, ECI 124Gemma 2 27B : Average price ($ per million tokens, log scale) $0.7, ECI 122Mistral Large : Average price ($ per million tokens, log scale) $3, ECI 122Mistral Nemo : Average price ($ per million tokens, log scale) $0.03, ECI 119Gemma 3 4B : Average price ($ per million tokens, log scale) $0.04, ECI 116Gemini 3.1 Flash Lite Preview : Average price ($ per million tokens, log scale) $0.6, ECI 144Gemini 3.1 Flash Lite PreviewClaude Opus 5.5
Each dot is a model rated by Epoch AI; Gemini 3.1 Flash Lite Preview is in orange. Higher = more capable, further left = cheaper. Open the LLM benchmark to filter.

For at least equal performance, Claude Haiku 5.5, GPT-6 Luna, and DeepSeek V4.1 Flash 1017 cost at least 30% less.

Features, privacy and local use

  • It supports tool calling, so it can drive an AI agent, call APIs or MCP servers.
  • It can answer in an enforced JSON format (structured outputs).
  • It is a reasoning model: it "thinks" before answering, which helps on hard tasks but uses more output tokens.

Its weights are not published: it can only be used through the Google API or a broker such as OpenRouter, which means sending your data off your machines.

Other Google models

Average price per million tokens (3 input for 1 output).
ModelReleasedAvg. priceECI
Gemini 4 ArgonOct 13, 2026$4166.5
Nano Banana 2.1Oct 6, 2026$3—
Gemini 3.8 FlashSep 2, 2026$1.65157.4
Gemini 3.7 FlashAug 13, 2026$1.5158
Gemini 3.6 FlashJul 21, 2026$1.5154.5
Gemini 3.5 Flash LiteJul 21, 2026$0.85144.9
Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)Jun 30, 2026$0.56—
Nano Banana 2 (Gemini 3.1 Flash Image)Jun 18, 2026$1.13—

Verdict

Gemini 3.1 Flash Lite Preview scores 55/100 on our grid, which only uses published figures.

Strengths

  • Very large context (1M tokens): long documents or a whole codebase fit in one request.
  • Agent-ready: tool calling and structured JSON output.

Avoid if…

  • your data must not leave your servers.

Alternatives

Comparisons and tools

Learn more

Frequently asked questions

How much does Gemini 3.1 Flash Lite Preview cost?

$0.25 per million input tokens and $1.5 per million output tokens on OpenRouter (checked October 25, 2026). A task with 2,000 tokens read and 500 written costs $0.0013.

How big is Gemini 3.1 Flash Lite Preview's context window?

1,048,576 tokens, roughly 786,000 words, with a maximum of 65,536 output tokens.

Can Gemini 3.1 Flash Lite Preview run locally?

No, Google does not publish its weights: it is API-only.

Is Gemini 3.1 Flash Lite Preview suitable for an AI agent?

Technically yes: it supports tool calling and structured outputs. Then check its cost for your volume.

How good is Gemini 3.1 Flash Lite Preview?

Epoch AI ranks it #87 of 125 rated models (ECI 144.3). On GPQA Diamond, it scores 81.8%.

What are the alternatives to Gemini 3.1 Flash Lite Preview?

The closest models on price and performance are MiniMax M3, GPT-5 Mini, and Qwen3.6 Flash.

Related articles

Neutrality. No affiliate links. No vendor pays to be reviewed, rated or ranked. Reviews are directed by François Chollet, neutral and factual, and every data point is dated and sourced. How we score