Skip to content
Rival
Models
CompareBest ForArena
Lab
Sign Up
Sign Up

We compare AI models for a living. On purpose. We chose this.

@rival_tips

Explore

  • Compare Models
  • All Models
  • Prompt Lab
  • Image Generation
  • Audio Comparison
  • Leaderboard
  • Challenges

Discover

  • Insights
  • AI Creators
  • AI Tools
  • The Graveyard

Developers

  • Developer Hub
  • MCP Server
  • .llmignore
  • Badges
  • Rival Datasets

Connect

  • Methodology
  • Sponsor
  • Partnerships
  • Privacy Policy
  • Terms
  • RSS Feed
© 2026 Rival · Built at hours no one should be awake, on hardware we don’t own
Rival
Models
CompareBest ForArena
Lab
Sign Up
Sign Up

We compare AI models for a living. On purpose. We chose this.

@rival_tips

Explore

  • Compare Models
  • All Models
  • Prompt Lab
  • Image Generation
  • Audio Comparison
  • Leaderboard
  • Challenges

Discover

  • Insights
  • AI Creators
  • AI Tools
  • The Graveyard

Developers

  • Developer Hub
  • MCP Server
  • .llmignore
  • Badges
  • Rival Datasets

Connect

  • Methodology
  • Sponsor
  • Partnerships
  • Privacy Policy
  • Terms
  • RSS Feed
© 2026 Rival · Built at hours no one should be awake, on hardware we don’t own
Rival
Models
CompareBest ForArena
Lab
Sign Up
Sign Up

Gemini 2.5 Flash Preview 05-20 (thinking) — AI Model Review

  1. Home
  2. Models
  3. Gemini 2.5 Flash Preview 05-20 (thinking)
Updated May 20, 2025
Share
Best for:Frontend DevelopmentCreative CodingUI Replication

Gemini 2.5 Flash Preview 05-20 (thinking) performance data on Rival is based on blind head-to-head community voting. Overall win rate: 53.3% across 45 duels. All vote data is part of Rival's open dataset of 21,000+ human preference judgments across 200+ AI models. Model responses are curated from 12 challenges.

Gemini 2.5 Flash Preview 05-20 (thinking)

Gemini 2.5 Flash Preview 05-20 (thinking)

Gemini:
Pro 1.0
1.5 Pro
2.0 Flash Thinking
2.0 Pro Experimental
2.5 Pro Experimental
2.5 Flash Preview 05-20 (thinking)

Gemini 2.5 Flash May 20th Checkpoint is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in "thinking" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. Note: This model is available in two variants: thinking and non-thinking. The output pricing varies significantly depending on whether the thinking capability is active. If you select the standard variant (without the ":thinking" suffix), the model will explicitly avoid generating thinking tokens. To utilize the thinking capability and receive thinking tokens, you must choose the ":thinking" variant, which will then incur the higher thinking-output pricing. Additionally, Gemini 2.5 Flash is configurable through the "max tokens for reasoning" parameter.

ConversationReasoningCode GenerationAnalysis
WebsiteOpenRouter
Feature this modelAdd badge to README
Provider
Google
Release Date
2025-05-20
Size
MEDIUM
Parameters
Not disclosed
Pricing
In: $0.15/1M
Out: $3.5/1M

API Access

Use Gemini 2.5 Flash Preview 05-20 (thinking) in your applications via the OpenRouter API. Copy the code below to get started.

import requests

response = requests.post(
"https://openrouter.ai/api/v1/chat/completions"    ,
    headers={
"Authorization""Bearer $OPENROUTER_API_KEY"        : ,
"Content-Type""application/json"        : 
    },
    json={
"model""google/gemini-2.5-flash-preview-05-20:thinking"        : ,
"messages""role""user""content""Hello!"        : [{: , : }]
    }
)
print(response.json())

Replace $OPENROUTER_API_KEY with your API key from openrouter.ai/keys

Model Insights

Model Responses

12 outputs from Gemini 2.5 Flash Preview 05-20 (thinking)

gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)Retro macOS UI Recreation
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)
Retro macOS UI Recreation
Run in Lab
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)Random Tailwind Webapp
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)
Random Tailwind Webapp
Run in Lab
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)Generate a Stand-Up Routine
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)

(Lights up on a stage. Comedian walks to center, adjusts mic, smiles warmly.)

"Alright, alright, settle down you crazy people! Thanks for having me. (Pause fo r applause/laughter)

You know, I was scrolling through Instagram earlier, and it's just a constant ba rrage of people 'living thei r best life,' isn't it? Someone's always 'grateful' for their 'morning yoga with a sunrise view' while simultaneously 'crushing their kale smoothie.' And you're just sitting there, in sweatpants, wondering if that kale smoothie would pair w ell with a donut. Because my best life looks suspiciously like their worst han gover.

And don't even get me started on autocorrect. It's like my phone has a personal vendetta against my dignity. I was trying to text my mom I was 'coming home' and it changed to 'I'm culmina ting home.' Like I'm some sort of obscure vegetable reaching its peak season. " Look Mom, I'm fully ripe! Eat me!"
(Pause for laughter)

Or trying to apologize for being late to a friend: 'Sorry, running a few minutes behind.' Autocorrect: 'Sorr y, running a few minutes behind schedule, you pathetic loser! (Mimes looking a t phone with horror) My phone is a cyberbully! I swear, one day it's gonna corre ct 'I love you' to 'I love gravy' and my relationship is just over.

(Pause)

And speaking of technology that actively hates us – who designed the self-checko ut machines? Every time I us e one, it's like a passive-aggressive game of 'guess the weight.' I bought one a pple. ONE. And it screams, in that robotic voice, 'UNEXPECTED ITEM IN BAGGING AR EA!' Sir, madam, it's an apple. My only item! Did you expect a pet hamster? A small, highly trained squirrel?

Then you have to wait for the teenager with the thousand-yard stare to come over and clear the ghost item. T hey always look like you just woke them up from a nap in a parallel dimension wh ere self-checkout never existed. And they just slowly press the button, no eye c ontact, like they're performing ancient magic. "Behold, the Bagging Area is clea r... for now, mortal."

(Pause, slightly exasperated face)

It makes you wonder, doesn't it? We're all supposed to be these highly functioni ng adults, navigating the di gital age, being productive members of society... but honestly, most days I feel like I'm just an overgrown toddler trying to assemble IKEA furniture using only a fork and the power of positive thinking.
And then I still go back to Instagram and share a picture of my perfectly assemb led – but wobbly – IKEA tabl e, captioning it: 'So blessed. So grateful.'

(Smile, gesture to audience)

You've been a great crowd! Thank you so much! I'm [Your Name], good night!"

(Walks off stage to applause)

Generate a Stand-Up Routine
Run in Lab
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)Surprise Animal SVG
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)
Surprise Animal SVG
Run in Lab
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)Math Misconception Test
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)

No, 9.11 is not greater than 9.9.

Here's how to compare them:

  • Compare the whole numbers: Both have 9.
  • Compare the first digit after the decimal point (the tenths place):
    • In 9.11, the digit is 1.
    • In 9.9, the digit is 9.

Since 1 is less than 9, 9.11 is less than 9.9.

You can also think of 9.9 as 9.90. Then comparing 9.11 and 9.90:

  • 9.11 (one hundred eleven)
  • 9.90 (nine hundred ninety)

Clearly, 9.11 is smaller.

Math Misconception Test
Run in Lab
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)Pokémon Battle UI Recreation
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)
Pokémon Battle UI Recreation
Run in Lab
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)World Map SVG
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)
World Map SVG
Run in Lab
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)Stochastic Consistency Test
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)

Here are 5 jokes:

  1. Why don't scientists trust atoms? Because they make up everything!
  2. What do you call a fake noodle? An impasta!
  3. Why did the scarecrow win an award? Because he was outstanding in his field!
  4. I told my wife she was drawing her eyebrows too high. She looked surprised.
  5. What do you call an alligator in a vest? An investigator!

All five jokes are unique. Each has a distinct setup, punchline, and humorous premise, even though some use similar mechanisms (like puns). They are not variations of the same core joke.

Stochastic Consistency Test
Run in Lab
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)SVG Layout Challenge
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)
SVG Layout Challenge
Run in Lab
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)Interactive Catan Board
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)
Interactive Catan Board
Run in Lab
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)Xbox Controller SVG Art
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)
Xbox Controller SVG Art
Run in Lab
Sponsored
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)Music App Landing Page
gemini-2.5-flash-preview-05-20-thinking logo
Gemini 2.5 Flash Preview 05-20 (thinking)
Music App Landing Page
Run in Lab

Related Models

Google: Gemini 3.1 Flash Lite Preview logo

Google: Gemini 3.1 Flash Lite Preview

Gemini 3.1 Flash Lite Preview is Google's high-efficiency model optimized for high-volume use cases. It outperforms Gemini 2.5 Flash Lite on overall quality and approaches Gemini 2.5 Flash performance across key capabilities. Improvements span audio input/ASR, RAG snippet ranking, translation, data extraction, and code completion. Supports full thinking levels (minimal, low, medium, high) for fine-grained cost/performance trade-offs. Delivers 2.5x faster Time to First Answer Token and 45% increased output speed compared to 2.5 Flash. Priced at half the cost of Gemini 3 Flash.

ConversationReasoningAnalysis+4 more
Gemini 3.1 Pro Preview logo

Gemini 3.1 Pro Preview

Gemini 3.1 Pro Preview is Google's frontier reasoning model, delivering enhanced software engineering performance, improved agentic reliability, and more efficient token usage across complex workflows. Building on the multimodal foundation of the Gemini 3 series, it combines high-precision reasoning across text, image, video, audio, and code with a 1M-token context window. The 3.1 update introduces measurable gains in SWE benchmarks and real-world coding environments, along with stronger autonomous task execution in structured domains such as finance and spreadsheet-based workflows. Designed for advanced development and agentic systems, it introduces a new medium thinking level to better balance cost, speed, and performance.

ConversationReasoningCode Generation+3 more
Gemini 2.5 Flash Lite Preview 06-17 logo

Gemini 2.5 Flash Lite Preview 06-17

Gemini 2.5 Flash Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, "thinking" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the Reasoning API parameter to selectively trade off cost for intelligence.

ConversationReasoningAnalysis+1 more
Gemini 2.5 Pro Preview 06-05 logo

Gemini 2.5 Pro Preview 06-05

Gemini 2.5 Pro is Google's state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs "thinking" capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities. Pricing: $1.25/M input tokens, $10/M output tokens, $5.16/K input images.

ConversationReasoningCode Generation+2 more
Gemini 2.5 Flash Preview 05-20 logo

Gemini 2.5 Flash Preview 05-20

Gemini 2.5 Flash May 20th Checkpoint is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in "thinking" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. Note: This model is available in two variants: thinking and non-thinking. The output pricing varies significantly depending on whether the thinking capability is active. If you select the standard variant (without the ":thinking" suffix), the model will explicitly avoid generating thinking tokens. To utilize the thinking capability and receive thinking tokens, you must choose the ":thinking" variant, which will then incur the higher thinking-output pricing. Additionally, Gemini 2.5 Flash is configurable through the "max tokens for reasoning" parameter.

ConversationReasoningCode Generation+1 more
Gemini 2.5 Flash Preview logo

Gemini 2.5 Flash Preview

Google's state-of-the-art workhorse model, designed for advanced reasoning, coding, mathematics, and scientific tasks. Features hybrid reasoning (thinking on/off) with configurable budgets, balancing quality, cost, and latency.

ConversationReasoningCode Generation+1 more

Prompt Lab

Don’t take our word for it.

Send your own prompt to Gemini 2.5 Flash Preview 05-20 (thinking) and 200+ others. See what comes back.

Keep exploring

COMPARE

Gemini 2.5 Flash Preview 05-20 (thinking) vs MiniMax M2.5

Real outputs compared side by side

RANKINGS

Best AI for Technical Analysis

Compare AI models on deep technical reasoning. Ranked across FLOP estimation,...

Compare Gemini 2.5 Flash Preview 05-20 (thinking)

Google: Gemini 3.1 Flash Lite PreviewNewer

We compare AI models for a living. On purpose. We chose this.

@rival_tips

Explore

  • Compare Models
  • All Models
  • Prompt Lab
  • Image Generation
  • Audio Comparison
  • Leaderboard
  • Challenges

Discover

  • Insights
  • AI Creators
  • AI Tools
  • The Graveyard

Developers

  • Developer Hub
  • MCP Server
  • .llmignore
  • Badges
  • Rival Datasets

Connect

  • Methodology
  • Sponsor
  • Partnerships
  • Privacy Policy
  • Terms
  • RSS Feed
© 2026 Rival · Built at hours no one should be awake, on hardware we don’t own
Nostalgia
Game Development
Interactive UI
Gemini 3.1 Pro PreviewNewer
Gemini 2.5 Flash Lite Preview 06-17Newer
Gemini 2.5 Pro Preview 06-05Newer
Gemini 2.5 Flash Preview 05-20google
Gemini 2.5 Flash PreviewOlder
Gemini 2.5 Flash Preview (thinking)Older
Gemini 2.5 Pro ExperimentalOlder

Alternatives to Gemini 2.5 Flash Preview 05-20 (thinking)

Gemini 2.5 Flash Preview 05-20 (thinking)’s competitors exist and they’ve been quietly putting in work. We thought you should know.

GPT-5.4 logo
GPT-5.4openai
Inception: Mercury 2 logo
Inception: Mercury 2inception
Qwen: Qwen3.5 35B A3B logo
Qwen: Qwen3.5 35B A3B
Claude Sonnet 4.6 logoRecraft V4 logo
Recraft V4recraft-ai
MiniMax M2.5 logo
MiniMax M2.5minimax
Z.ai: GLM 5 logo
qwen
Claude Sonnet 4.6anthropic
Z.ai: GLM 5zhipu