NOUMENOpen app
NOUMEN

Google · Efficient

Gemini 3.5 Flash

Added May 19, 2026

Compare models

Choose up to 3 models to compare.0 of 3 chosen

Loading models…

Use in the app

Google's speed-focused Gemini Flash model for frontier multimodal intelligence across text, images, audio, video, PDFs, and code. Built for agentic coding, reliable tool use, structured outputs, and long-context workflows. Audio input costs $3.00 per million audio tokens.

  • Reasoning
  • Tools
  • Vision
  • Structured output
  • Audio input
  • Video input
Input
Text, Images, Video, Audio
Output
Text
Reasoning effort
minimal · low · medium · high
Discovery tier
Efficient

Pricing

USD per 1M tokens
Input
$1.50
Output
$9.00
Cache read
$0.15
Cache write
$0.0833

Reference prices as published by the model provider. API example

Specifications

Context window
1M
Max output
65.5K
Average output (7 days)
322 tokens
Longer than 35% of 473 models

Side by side

Compare models

Put Gemini 3.5 Flash next to up to three other models. The comparison stays in the link, so you can share it.

Compare models

Choose up to 3 models to compare.0 of 3 chosen

Loading models…

TryGPT 5.3 CodexClaude 4.6 Opus ThinkingMuse Spark 1.1All 3

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

Intelligence Index

32.6

Tested at high reasoning effort

Better than 76% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    70.1

    Composite of the coding evaluations

    Better than 77% of 168 scored models

Agentic work

  • AutomationBench-AA

    42.1%

    Workflow automation, penalized for guardrail violations

    Better than 58% of 173 scored models

  • AutomationBench-AA tasks completed

    13.5%

    Workflows finished end to end without a guardrail violation

    Better than 56% of 173 scored models

  • AA-Briefcasev1.1

    874 Elo

    Agentic knowledge work, rated head to head

    Better than 36% of 166 scored models

  • GDPval-AAv2.1

    1,202 Elo

    Economically valuable occupational tasks, rated head to head

    Better than 51% of 187 scored models

Document reasoning

  • GDP.pdf

    19.8%

    Professional reasoning over PDF documents, all-pass rate

    Better than 76% of 171 scored models

  • AA-LCRv1.1

    73.3%

    Reasoning across long documents

    Better than 55% of 309 scored models

  • MLCR-AA

    10.0%

    Long-context reasoning over medical records

    Better than 33% of 121 scored models

Reasoning and knowledge

  • Humanity’s Last Exam

    42.7%

    Expert-level questions across academic fields

    Better than 84% of 323 scored models

  • GPQA Diamond

    92.2%

    Graduate-level science questions

    Better than 87% of 291 scored models

  • IFBench

    76.3%

    Following precise output instructions

    Better than 92% of 215 scored models

Coding

  • Terminal-Benchv4.0

    6.6%

    Practical software engineering in a terminal

    Better than 57% of 171 scored models

  • SciCode

    53.9%

    Scientific computing in Python

    Better than 68% of 174 scored models

Legacy evaluations

  • Terminal-Bench Hard

    40.9%

    Earlier terminal task suite

    Better than 77% of 212 scored models

  • τ²-Bench Telecom

    95.3%

    Tool use in a customer-support simulation

    Better than 86% of 214 scored models

  • AA-LCR (unversioned)

    73.3%

    Long-context reasoning, earlier grading

    Better than 55% of 309 scored models

Scores from Artificial Analysis, updated Oct 26, 2026. Gemini 3.5 Flash was tested at high reasoning effort; compare scores measured at the same effort.Artificial Analysis

Live routing data

Performance and reliability

Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.

Throughput
264.0 tok/s
Faster than 91% of 256 models
Time to first token
4.8 s
Quicker than 30% of 256 models
Prompt cache hits
13.2%
About $1.32 per 1M input tokens with caching
Request success (24h)
98.0%
Reliable

Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.

Hourly request success · last 72 hoursLowest hour 70.0%
  • Healthy (98% or more)
  • Degraded (90–98%)
  • Down (under 90%)
  • No requests

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart
curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemini-3.5-flash",
    "reasoning_effort": "medium",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'
import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="google/gemini-3.5-flash",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Google

Other Google models, starting with the same family.