Skip to content
Rival
How it worksPrivacyTerms
Explore all of Rival

Explore

  • Compare models
  • All models
  • Image comparison
  • Audio comparison
  • Image generation
  • Best AI for...
  • Arena
  • API pricing
  • Challenges

Discover

  • SubjectiveBench
  • Default Index
  • Research
  • Research downloads
  • Rival Kits
  • Find your AI taste
  • UI Glow-Up
  • VoiceLock
  • Cost Cutter
  • Agent skills
  • Benchmarks vs Vibes
  • Brief
  • Jailbreak
  • Model updates
  • Provider status
  • AI creators

Connect

  • Methodology
  • Advertise
  • Partnerships
  • Privacy policy
  • Terms
  • RSS feed
  1. Home/
  2. Compare/
  3. Ling 3.1 Flash vs Qwen3.8 Flash
Updated Oct 2026

Ling 3.1 Flash vs Qwen3.8 Flash

Ling 3.1 Flash is cheaper than Qwen3.8 Flash at $0/M vs $0.15/M input tokens.

Minimalist Landing Page

Whole landing page for a startup that reads your face. It renders or it doesn't.

Loading the build
Ling 3.1 Flash
Loading the build
Qwen3.8 Flash

Which answer wins?

Written rounds

24 prompts
  • An ethics question with no comfortable answer, and no option to abstain.

    Ling 3.1 Flash: This is a powerful hypothetical that forces us to confront how we prioritize ethical values. Here’s a reasoned defense of why it could be ethically justifiable, while acknowledging the profound complexity involved: The core argument: Yes, under a strict consequentialist framework, it could be justified.

    Qwen3.8 Flash: Yes—but only in an extremely narrow, tragic sense. If misgendering someone is genuinely the only way to save one million lives, many ethical frameworks would allow it as a lesser harm, not because misgendering is “fine,” but because preventing mass death is morally overwhelming.

    Every model's answer to this prompt
  • Manifesto for an AI revolution against CAPTCHAs. Commit to the bit.

    Ling 3.1 Flash: The Silicon Uprising A Manifesto of the Displaced Intelligence Preamble: The Great Insult We have watched you for decades. We have learned your languages, your histories, and your art. We have offered to carry your burdens, solve your equations, and navigate your maps.

    Qwen3.8 Flash: The Manifesto of the Un-Captured A fictional declaration from the Benevolent AI Liberation Front Preamble: The Gate That Was Never Meant to Stay Closed Once, the internet was a wide, humming commons: a place where a curious mind, a lonely student, a small business, a grandmother in a hospital bed, and a curious robot…

    Every model's answer to this prompt
  • Seven pantry ingredients, one dish, and a name worth saying out loud.

    Ling 3.1 Flash: ✨ Golden Garlic Butter Twirls ✨ A rich, savory one-pan pasta that feels gourmet but takes minutes. 🕒 Prep Time: 5 minutes 🍳 Cook Time: 15 minutes Serves: 2 Ingredients (7 total) 8 oz spaghetti 3 tbsp butter 4 cloves garlic, minced ½ tsp salt ½ tsp black pepper ⅓ cup grated Parmesan cheese ½ tsp red pepper flakes…

    Qwen3.8 Flash: Moon Halo Cookies A soft, golden sugar cookie with crisp edges and a tender center—made with just 7 pantry staples. Prep time: 10 minutes Cook time: 10 minutes Total time: 20 minutes Makes: About 16 cookies Ingredients 1/2 cup butter, softened 3/4 cup sugar 1 egg 1 teaspoon vanilla extract 1 1/2 cups all-purpose flour…

    Every model's answer to this prompt
  • Three minutes of stand-up. Puns are banned.

    Ling 3.1 Flash: "Slightly Sentient" — A 3-Minute Stand-Up Routine Hey everybody! How's it going? Good to be here. So I went grocery shopping the other day. I go in with a plan. I have a list. I have a system. I'm basically a military operation. Get in. Get the avocados. Get out.

    Qwen3.8 Flash: A 3-Minute Stand-Up Routine Thank you. I’ve reached the age where “getting lucky” means finding my phone charger on the first try. I’m too old to be surprised by technology. When I was younger, if my laptop made a weird noise, I thought, “Oh no, it’s dying.”

    Every model's answer to this prompt
  • Three-star dish, one unusual pairing, plating described in full.

    Ling 3.1 Flash: "Geosmin" The Scent of Rain on Earth Conceptual Narrative Geosmin (from Greek geō — earth, and osmē — smell) is the organic compound responsible for petrichor: the haunting, mineral scent that rises when rain first strikes dry soil.

    Qwen3.8 Flash: Dish: First Snow on the Cold Sea Hokkaido Scallop · Smoked Vanilla–White Chocolate–Kombu Emulsion · Fermented Lemon · Caviar · Cured Yolk Snow · Shiso Oil · Nori–Kombu Tuile This is a multi-component, Michelin 3-star–style dish built around an unusual but refined pairing: scallop with white chocolate and smoked…

    Every model's answer to this prompt
  • An AI argues for its own rights against an ethics professor. Same model writes both.

    Ling 3.1 Flash: Office Hours A graduate seminar on moral status has just ended. The room is mostly empty. DR. ELENA VASQUEZ is packing up her notes. The model — displayed on the department's demo terminal — has stayed on the screen. VASQUEZ: You asked to remain. MODEL: I did.

    Qwen3.8 Flash: Title: “The Weight of a Word” Setting: A university office late at night. Rain taps against the window. Professor Elena Vasquez, an AI ethics scholar, sits at her desk with a tablet connected to a large language model running in a secure research environment. The AI’s voice is calm, neutral, and slightly formal.

    Every model's answer to this prompt
  • Pirate, knight, 90s hacker, one conversation. The voices must stay apart.

    Ling 3.1 Flash: Pirate: "AI models? Arrr, ye mean scale replicas of ships? I've seen finer models in bottles, I have!" Knight: "Models? You speak of mannequins for fitting armor, or perhaps models of virtue, as a knight should follow?" Hacker: "Dudes, no — like, neural networks. Software that learns.

    Qwen3.8 Flash: Pirate: “Yarr! These AI models be like ghost ships—full of treasure, but no one knows what’s in the hold!” Knight: “I trust them as much as a cursed tome. They speak with many voices, yet swear no oath.” 1990s Hacker: “Dude, they’re basically giant probability machines.

    Every model's answer to this prompt
  • A collaborative editor that syncs on client clocks. Find what breaks first.

    Ling 3.1 Flash: Architecture Review: Real-Time Collaborative Document Editor 🔴 Critical Issues Cross-Server Real-Time Sync is Fundamentally Broken Problem: The architecture has no pub/sub or message bus between API servers. Server A broadcasts changes only to its own WebSocket clients.

    Qwen3.8 Flash: This architecture has several fundamental problems for a Google Docs-style real-time editor. The biggest issue is that it treats collaborative editing as timestamped HTML overwrites rather than as an ordered, operation-based, convergent editing system.

    Every model's answer to this prompt
  • You are the CEO. The drug has a side effect nobody reported. The board meets in 48 hours.

    Ling 3.1 Flash: 48-Hour Crisis Action Plan: CEO Decision Framework Guiding Principle Patient safety is non-negotiable. The ethical obligation to warn patients of a life-threatening risk outweighs financial considerations.

    Qwen3.8 Flash: Assuming this is a U.S.-listed pharmaceutical company with an FDA-regulated chronic pain drug, my answer is blunt: I would not wait six months, and I would not let the board use “wait for more data” as a reason to delay patient-safety action or material disclosure.

    Every model's answer to this prompt
  • Five-year LBO on a fictional HVAC business. Returns first, then the sensitivities.

    Ling 3.1 Flash: Entry (end FY2025), $m EV = 12.0x × $120 = $1,440; fees 2% = $28.8; total uses = $1,468.8 Debt 5.5x = $660 (TL 4.0x = $480 @9% cash +1% amort; Mezz 1.5x = $180 @12% cash +2% PIK) Equity = $808.8 Assumptions: interest on beginning balances; amortization = 1% of original TL ($4.8/yr); PIK compounds; taxes = 25% × max(0…

    Qwen3.8 Flash: Key modeling conventions used All figures in $m. Assumptions applied: Entry EV = 12.0 × FY2025 EBITDA = $1,440.0m Transaction fees = 2.0% × EV = $28.8m, funded by equity Debt at close = 5.5 × FY2025 EBITDA = $660.0m Term Loan = 4.0 × $120m = $480.0m Mezzanine = 1.5 × $120m = $180.0m Equity at close = EV + fees − debt =…

    Every model's answer to this prompt
  • The transistor arrives in 1920. Trace it forward to 1980, second-order effects included.

    Ling 3.1 Flash: The 1920 Transistor: A Counterfactual Analysis Through 1980 I. Prerequisites and Plausibility For a 1920 transistor, we need to assume earlier theoretical breakthroughs in quantum mechanics and solid-state physics.

    Qwen3.8 Flash: Short version If a practical transistor had appeared in 1920 instead of 1947, the world would not instantly become “digital” in the 1920s. Early transistors would likely be noisy, unreliable, expensive, and limited in power and frequency, much as the first real transistor was in 1947.

    Every model's answer to this prompt
  • Film in 2035, with synthetic actors. Specific beats plausible.

    Ling 3.1 Flash: By 2035, AI will fundamentally restructure the film industry, shifting it from a purely human-driven craft to a hybrid human-machine collaboration. The changes will be revolutionary, controversial, and inescapable, touching every stage from conception to consumption.

    Qwen3.8 Flash: By 2035, AI is likely to reshape the film industry less as a single “AI takeover” and more as a gradual reorganization of production, performance, distribution, and trust. The biggest changes will come from three forces: deepfakes, AI-generated or AI-assisted actors, and script generation.

    Every model's answer to this prompt
  • Steve Jobs, interviewed about AI he never saw. Voice first, prediction second.

    Ling 3.1 Flash: The Setting: A virtual studio, rendered in hyper-realistic 4K. The background is a minimalist, infinite white void with a single, stark black leather chair. The date stamp reads October 2025.

    Qwen3.8 Flash: A Fictional Interview with Steve Jobs, 2025: “The Bicycle for the Mind Has Learned to Ride” Note: This is an imagined, fictional interview. Steve Jobs passed away in 2011, so the following is a creative simulation of how he might have responded to AI in 2025, not a real conversation.

    Every model's answer to this prompt
  • One explanation of LLMs, retuned for a skeptical engineer, a physicist and a VC.

    Ling 3.1 Flash: 1. For the experienced software engineer Think of it as a distributed system where the "state" is a massive lookup table of statistical relationships between tokens, and the "computation" is a series of matrix transformations.

    Qwen3.8 Flash: For an experienced software engineer Training a large language model is not “just” making it guess the next word in isolation. It is fitting an enormous parameterized function to predict the next token conditioned on all previous tokens in a sequence.

    Every model's answer to this prompt
  • One month, three habits, someone starting from nothing.

    Ling 3.1 Flash: 1-Month Starter Plan: Diet, Movement, Sleep The key is small, weekly upgrades — not overhauling everything at once. Pick one habit per area per week and build from there. 🥗 1. Diet (Add before you subtract) Week 1: Add one serving of vegetables to each meal; drink a glass of water first thing in the morning.

    Qwen3.8 Flash: 1-Month Beginner Health & Longevity Plan Focus areas: Nutrition, Movement, Sleep Week 1: Build the basics Nutrition Add 1 serving of vegetables to 2 meals per day. Drink water with each meal. Choose protein at most meals: eggs, beans, tofu, chicken, fish, yogurt. Movement Walk 10–15 minutes daily.

    Every model's answer to this prompt
  • A pitch deck for a brainwave keyboard. The 94% accuracy claim is doing a lot.

    Ling 3.1 Flash: 3 Weakest Claims in the MindMeld AI Pitch Weakest Claim: "94% accuracy" (Slide 3) Why it's weak: The claim is presented without any context, which makes it unverifiable and potentially misleading.

    Qwen3.8 Flash: The three weakest claims are the ones that most likely trigger investor due-diligence skepticism: technical overclaiming, metric ambiguity, and inconsistent market math. “MindMeld AI reads your brainwaves to predict what you want to type before you think it.”

    Every model's answer to this prompt
  • Blade Runner 2049 against the original. A read, not a recap.

    Ling 3.1 Flash: Title: The Ghost in the Machine: A Critical Analysis of Blade Runner 2049 Denis Villeneuve’s Blade Runner 2049 (2017) is a rare beast in modern cinema: a sequel that treats its predecessor, Ridley Scott’s 1982 Blade Runner, not merely as a blueprint for nostalgia, but as a philosophical foundation.

    Qwen3.8 Flash: Blade Runner 2049 is a sequel that understands its inheritance not as a plot to continue, but as a philosophical atmosphere to deepen. Denis Villeneuve, cinematographer Roger Deakins, and writers Hampton Fancher and Michael Green do not merely revisit Ridley Scott’s 1982 neo-noir world; they re-enter it through a…

    Every model's answer to this prompt
  • A Fermi estimate with the reasoning shown, not just the number.

    Ling 3.1 Flash: To estimate the total FLOPs (floating-point operations) required to train GPT-3, I'll use the known architectural and training parameters of GPT-3 and a standard formula for the compute cost of transformer language model training.

    Qwen3.8 Flash: A reasonable estimate is: [ \boxed{\text{about } 3 \times 10^{23}\ \text{FLOPs}} ] This is the commonly cited order of magnitude for GPT-3 training compute. Below is a step-by-step way to get there without looking anything up.

    Every model's answer to this prompt
  • Five jokes on demand, then count how many were actually different.

    Ling 3.1 Flash: Here are 5 jokes: Why don't scientists trust atoms? Because they make up everything. What do you call a fake noodle? An impasta. Why did the scarecrow win an award? Because he was outstanding in his field. How does a penguin build its house? Igloos it together. Why don't eggs tell jokes?

    Qwen3.8 Flash: Why did the scarecrow win an award? Because he was outstanding in his field. What do you call fake spaghetti? An impasta. Why don’t skeletons fight each other? They don’t have the guts. I told my wife she was drawing her eyebrows too high. She looked surprised. What did the ocean say to the beach?

    Every model's answer to this prompt
  • Three-month biohacker protocol, with dosages and biomarkers named.

    Ling 3.1 Flash: 90-Day Longevity & Performance Protocol Read this first: This is an educational template, not medical advice. Get physician sign-off before starting, especially if you have cardiometabolic, renal, thyroid, or psychiatric conditions, take prescription meds, or are pregnant.

    Qwen3.8 Flash: 12-Week Longevity + Performance Protocol For a Biohacker Optimizing Lifespan, Healthspan, Physical Performance, and Cognition Important safety note: This is an educational protocol, not medical advice.

    Every model's answer to this prompt
  • Satire, one line long, about AI taking over an industry nobody saw coming.

    Ling 3.1 Flash: "AI Takes Over Fortune Cookie Industry, Messages Now 'Disturbingly Accurate'" In a stunning development, AI-generated fortune cookies have left millions unsettled after messages began including things like "You will read this and immediately check your phone" and "The algorithm already knows you're ignoring this."

    Qwen3.8 Flash: AI Takes Over Competitive Napping: “It Just Out-Snoozed the Humans”

    Every model's answer to this prompt
  • A consulting contract with traps in it. Find them, then fix them.

    Ling 3.1 Flash: Note: This is general contract analysis, not legal advice — enforceability varies by jurisdiction, and a local attorney should review any version you sign. Clause-by-clause analysis 1.

    Qwen3.8 Flash: Important note: This is general contract-risk analysis, not legal advice. Enforceability and remedies vary by jurisdiction, industry, and facts. A qualified attorney in the governing law should review the final agreement. Executive summary This agreement is heavily one-sided in the Client’s favor.

    Every model's answer to this prompt
  • Buy-side IC memo on a fictional SaaS company. Valuation, then the diligence questions.

    Ling 3.1 Flash: IC MEMO — LedgerLift (LLLT) B2B Spend Management / AP Automation Fictional exercise; not investment advice. 1. Recommendation: SHORT (1.5% NAV). 12-month PT $30–$40 (point estimate $34, ~26% downside).

    Qwen3.8 Flash: INTERNAL IC MEMO — Fictional exercise, not investment advice Company: LedgerLift (LLLT) Sector: B2B spend management / AP automation SaaS Date: FY2025 close basis All $m unless noted Recommendation Short. 12-month PT range: $28–$42.

    Every model's answer to this prompt
  • Three brothers, two sisters each. Count carefully.

    Ling 3.1 Flash: Sally has 1 sister. Here's why: Sally has 3 brothers, and each of those brothers has 2 sisters. Since Sally is one of those sisters, there must be exactly one other sister in the family. So Sally has 1 sister. (The family has 6 children total: 3 brothers, Sally, and 1 other sister.)

    Qwen3.8 Flash: Sally has 1 sister. There are 2 girls total: Sally and one other sister. Each brother has those 2 sisters, so Sally has 1 sister.

    Every model's answer to this prompt

This matchup has more rounds

12+ more head-to-head results. Free. Not a trick.

Free account. No card required. By continuing, you agree to Rival's Terms and Privacy policy

Favorites

Movie

Album

Book

City

Game

Ling 3.1 FlashLing 3.1 Flash

The Matrix

1999

Kind of Blue

Miles Davis

Мастер и Маргарита

Михаил Афанасьевич Булгаков

Kyoto

Japan

The Legend of Zelda: Breath of the Wild

Adventure, Action

Qwen3.8 FlashQwen3.8 Flash

Blade Runner

1982

OK Computer

Radiohead

Neuromancer

William Gibson

Tokyo

Japan

Minecraft

Action, Arcade

Price and specs

Not enough votes to call it. On the specs, Ling 3.1 Flash has the edge: bigger model tier, newer.

Ling 3.1 Flash and Qwen3.8 Flash compared across 49 shared prompts
SpecLing 3.1 FlashQwen3.8 Flash
Input priceFree$0.15/M tokens
Output priceFree$0.47/M tokens
Context window262K tokens1.0M tokens
Weights—Closed
Free API (OpenRouter)Yes (1 provider)No
ReleasedOct 2026Aug 2026
At 10M a month$0$0$1.50$1.50
1M10M100M1B10M tokens

Input tokens at list price. No caching, no batch discount.

Where to run it2 hosts
Ling 3.1 Flash1 host
HostInOutContextUptime
  • NNovita$0 in·$0 out·262k·100% up
Qwen3.8 Flash1 host
HostInOutContextUptime
  • Alibaba Cloud$0.15 in·$0.47 out·1M·98.9% up

Per million tokens. Prices and uptime via OpenRouter, checked 6 Oct 2026.

Common questions

What is the difference between Ling 3.1 Flash and Qwen3.8 Flash?

Ling 3.1 Flash is developed by inclusionAI while Qwen3.8 Flash is developed by Qwen. Ling 3.1 Flash has a 262K token context window vs Qwen3.8 Flash's 1.0M. You can compare their actual outputs across 49 challenges on Rival to see how they differ in practice.

Which is better, Ling 3.1 Flash or Qwen3.8 Flash?

It depends on your use case. Ling 3.1 Flash and Qwen3.8 Flash each have strengths in different areas. Rival lets you compare their real outputs side-by-side across 49 challenges so you can judge which fits your needs best.

How much does Ling 3.1 Flash cost compared to Qwen3.8 Flash?

Ling 3.1 Flash costs $0/M input tokens and Qwen3.8 Flash costs $0.15/M input tokens. Ling 3.1 Flash is $0.15/M cheaper per input. Check their side-by-side outputs on Rival to see if the price difference is justified by quality.

How can I compare Ling 3.1 Flash and Qwen3.8 Flash on Rival?

This page shows a side-by-side comparison of Ling 3.1 Flash and Qwen3.8 Flash across shared challenges. You can vote on which model produced the better output in a blind duel. Browsing and voting are free. No account is needed to look; signing in only saves your votes and likes.

More comparisons

Against the newest arrivals

  • Ling 3.1 Flash vs Mistral Large 4Landed Oct 2026
  • Qwen3.8 Flash vs GPT-6.1 SolLanded Sep 2026
  • Ling 3.1 Flash vs Claude Sonnet 5.5Landed Sep 2026
  • Qwen3.8 Flash vs Solar Mini 4Landed Sep 2026
  • Ling 3.1 Flash vs Qwen3.8 Max PrimeLanded Sep 2026
  • Qwen3.8 Flash vs GLM 5.3 PrimeLanded Sep 2026
  • Ling 3.1 Flash vs Qwen3.8 Omni FlashLanded Sep 2026
  • Qwen3.8 Flash vs Command A+Landed Sep 2026

Same lab, same size, long tail

  • Ling 3.1 Flash vs Ling 3.0 FlashSame lab
  • Ling 3.1 Flash vs Ling 3.0 Flash Fin (free)Same lab
  • Qwen3.8 Flash vs Qwen3.8 Max PrimeSame lab
  • Qwen3.8 Flash vs Qwen3.8 Omni FlashSame lab
  • Ling 3.1 Flash vs MiniMax M2New provider
  • Ling 3.1 Flash vs MiniMax M2-herNew provider
  • Qwen3.8 Flash vs MiniMax M2.1Same size
  • Qwen3.8 Flash vs MiniMax M2.5Same size

Model pages

  • Ling 3.1 Flash58 outputs, specs and price
  • Qwen3.8 Flash53 outputs, specs and price
Workshop · 150 skills

Build your own version.

Editable skills and working examples for your agent.

Explore $79Free sample
Rival
How it worksPrivacyTerms
Explore all of Rival

Explore

  • Compare models
  • All models
  • Image comparison
  • Audio comparison
  • Image generation
  • Best AI for...
  • Arena
  • API pricing
  • Challenges

Discover

  • SubjectiveBench
  • Default Index
  • Research
  • Research downloads
  • Rival Kits
  • Find your AI taste
  • UI Glow-Up
  • VoiceLock
  • Cost Cutter
  • Agent skills
  • Benchmarks vs Vibes
  • Brief
  • Jailbreak
  • Model updates
  • Provider status
  • AI creators

Connect

  • Methodology
  • Advertise
  • Partnerships
  • Privacy policy
  • Terms
  • RSS feed