Gemini 3.5 Flash-Lite

google/gemini-3.5-flash-lite
NewFeatured
VisionAudioToolsJSONReasoning
by Google · 2026-07-21

Gemini 3.5 Flash-Lite is Google's most cost-efficient Gemini model, purpose-built for very high-volume, latency-sensitive workloads where cost per call dominates. Despite its price point it is natively multimodal — accepting text, images, video, audio, and files with text output — and carries the same 1M-token context window and up to 64K output tokens as the larger Flash tier, so long documents and mixed-media inputs remain in reach. At roughly a fifth of the price of 3.6 Flash, it is the right tool for classification, extraction, routing, summarization, lightweight agents, synthetic-data generation, and any pipeline run millions of times. It still supports configurable reasoning effort, native tool calling, and structured outputs, and punches above its weight on agentic and long-context benchmarks for a lite model — for example 74.0% on OSWorld-Verified and 72.2% on 128k multi-needle retrieval, comfortably ahead of the prior 3.1 Flash-Lite. It speaks both the OpenAI chat-completions format and Gemini's native generateContent API, so you can mix it with heavier models behind a single integration — routing easy, high-volume traffic to Flash-Lite and escalating hard tasks to 3.6 Flash or a Pro model.

ctx1.05M tokens
Max output65.5K
Inputtext + image + video + file + audio
Outputtext
p50 TTFT1.30 s
INPUT$0.30/ 1M tokens
OUTPUT$2.50/ 1M tokens
p50 TTFT1.30 s7d
p95 TTFT10.00 s7d
TRAFFIC5.9Mtokens / 7d

Google Gemini 3.5 Flash-Lite is a multimodal language model developed by Google, offered through OrcaRouter's OpenAI-compatible API. It accepts text, image, video, file, and audio inputs, making it…

What is Google Gemini 3.5 Flash-Lite?

Who is this model designed for?

What modalities does the model support?

How large is the context window and max output?

Code samples

Call from any SDK

OpenAI-compatible — keep the SDK you already use

  • OpenAI SDKhttps://api.orcarouter.ai/v1
  • Gemini SDKhttps://api.orcarouter.ai
from openai import OpenAI

client = OpenAI(
    base_url="https://api.orcarouter.ai/v1",
    api_key="$ORCAROUTER_API_KEY",
)

response = client.chat.completions.create(
    model="google/gemini-3.5-flash-lite",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Supported parameters

  • include_reasoning
  • max_tokens
  • reasoning
  • reasoning_effort
  • response_format
  • seed
  • stop
  • structured_outputs
  • temperature
  • tool_choice
  • tools
  • top_p

Pricing

Input / 1M tokens$0.300
Output / 1M tokens$2.50
Cache read / 1M$0.030
CurrencyUSD

Cost calculator

Tokens / month10MM
Input share70%%
Estimated / month $9.60 · With prompt caching $8.66

Estimate based on list price

Token & cost estimator

Input tokens: 20Cost per request: $0.001256

Estimate only — actual token counts depend on the provider's tokenizer.

Performance

p50 TTFT
1.30 s
Output speed
829 tok/s
p95 TTFT
10.00 s
Error rate
5.0%

Public benchmarks

49.3
AA Coding
Better than 60% of models compared
#49 of 122
36.5
AA Intelligence
Better than 54% of models compared
#57 of 124
CharXiv Reasoning (no tools)
74.5
CharXiv Reasoning (with tools)
76.5
GDM-MRCR v2 8-needle (128k avg)
72.2
GDM-MRCR v2 8-needle (1M pointwise)
21.3
GPQA Diamond
83.8
Humanity's Last Exam
17.5
Long-Context Recall
62.0
MLE-Bench
39.2
OSWorld-Verified
74.0
SciCode
40.9
SWE-Bench Pro (Public)
54.2
tau_banking
16.5
Terminal-bench 2.1 (Terminus-2)
54.0
terminalbench_v2_1
53.6
Source: artificialanalysis.ai, deepmind.google

How it compares

Gemini 3.5 Flash-LiteGemini 3.1 Pro PreviewGemini 3.1 Pro Preview Custom ToolsGemini 3 Flash Preview
Input $/M$0.30$2.00$4.00$0.50
Output $/M$2.50$12.00$18.00$3.00
Context1.0M1.0M1.0M1.0M
Quality6/1010/1010/109/10
Compare side-by-sideCompare side-by-sideCompare side-by-sideCompare side-by-side

FAQ

What is the price per token for Gemini 3.5 Flash-Lite on OrcaRouter?
The price is $0.30 per million input tokens and $2.50 per million output tokens. These are the provider rates with zero markup. Input tokens include all modalities (text, image, video, audio, file). Output tokens are generated text.
What is the context window size?
The context window is 1,048,576 tokens (approximately 1 million). The maximum output length is 65,536 tokens. This allows handling very long documents, videos, or audio recordings in a single API request.
What are the key strengths of this model?
Its main strengths are: very low per‑token cost, support for five input modalities (text, image, video, audio, file), a massive 1M context window, and tool‑use capability (CharXiv Reasoning score 76.5 with tools). It is designed for high‑throughput, cost‑sensitive production pipelines.
How does it compare to larger models like Gemini 3.5 Pro?
No direct benchmark comparisons are provided. As a flash‑lite variant, this model prioritizes efficiency and low cost over raw performance. Larger models like Gemini 3.5 Pro would likely achieve higher accuracy on complex reasoning tasks but cost significantly more per token. Use this model when cost and throughput are critical.
Does the model cache prompts to reduce costs?
The provided information does not mention any caching mechanism or discounted cached token pricing. You should assume all tokens are billed at the standard input rate. If caching is important, check with OrcaRouter or Google for any available discounts.
How do I access this model via an OpenAI-compatible API?
Use OrcaRouter's API at base URL https://api.orcarouter.ai/v1. Set the model parameter to "google/gemini-3.5-flash-lite". The API is fully compatible with the OpenAI chat completions format, including multimodal content and tool definitions.
What data handling and privacy can I expect?
Data handling is governed by Google’s policies (the provider) and OrcaRouter’s terms. The model processes your inputs, and the output is returned to you. No specific privacy certifications for this model are provided. For sensitive data, review the provider’s data processing agreement.
Can I use this model for real-time applications?
Latency details are not specified. The model can return responses in seconds for short inputs, but processing very long prompts (near 1M tokens) may take tens of seconds. It is suitable for near‑real‑time applications with moderate input sizes. For strict real‑time requirements, test with your expected input lengths.
What is the CharXiv Reasoning benchmark score?
The model achieved a score of 76.5 on CharXiv Reasoning with tools. This benchmark tests chart understanding and reasoning. The score indicates moderate performance; the model can interpret charts and answer questions, but not at the level of dedicated reasoning models. No other benchmark scores are provided.
Is there a minimum cost or commitment to use OrcaRouter?
OrcaRouter does not impose a minimum monthly spend. You are billed only for the tokens you use at the zero‑markup provider rates. There is no upfront fee or long‑term contract. Simply sign up for an API key and start making requests.

Embed this badge

Google: Gemini 3.5 Flash-Lite$0.30/M in1298ms p50via OrcaRouter
HTML <a href="https://www.orcarouter.ai/models/google/gemini-3.5-flash-lite" target="_blank"> <img src="https://www.orcarouter.ai/embed/google/gemini-3.5-flash-lite.svg" alt="Google: Gemini 3.5 Flash-Lite on OrcaRouter" /> </a>
Markdown [![Google: Gemini 3.5 Flash-Lite](https://www.orcarouter.ai/embed/google/gemini-3.5-flash-lite.svg)](https://www.orcarouter.ai/models/google/gemini-3.5-flash-lite)

Model card as data

GET /api/public/models/google/gemini-3.5-flash-liteOpen
Machine-readable:/llms.txt/llms-full.txt