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. Qwen3 Next 80B A3B Instruct vs Qwen3.5 Plus 2026-04-20
Updated Apr 2026

Qwen3 Next 80B A3B Instruct vs Qwen3.5 Plus 2026-04-20

Qwen3 Next 80B A3B Instruct is cheaper than Qwen3.5 Plus 2026-04-20 at $0.15/M vs $0.4/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
Qwen3 Next 80B A3B Instruct
Loading the build
Qwen3.5 Plus 2026-04-20

Which answer wins?

Written rounds

24 prompts
  • Pirate, knight, 90s hacker, one conversation. The voices must stay apart.

    Qwen3 Next 80B A3B Instruct: [Scene: A foggy dockside tavern. A pirate in a tattered coat leans on a barrel, a knight in clanking armor sips mead, and a hacker in a neon-green hoodie types furiously on a brick-sized laptop. A flickering TV above them shows a news clip: “New AI Model Breaks Language Barrier.”] Pirate (slamming tankard): “AI model?

    Qwen3.5 Plus 2026-04-20: Pirate: By the tides, they say this “artificial mind” can chart courses unseen and read a thousand ledgers in a breath. Sounds like a navigator that never sleeps and asks no questions. Knight: A curious construct. But can it be sworn to a code? Steel and oath bind a warrior; what binds this… model?

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

    Qwen3 Next 80B A3B Instruct: Recipe Name: Golden Honey-Sesame Toast Bites A crunchy, sweet-savory snack that’s warm, glossy, and addictive—perfect for breakfast, dessert, or a quick treat. Ingredients (7 pantry staples): Bread (white or whole wheat) — 4 slices Butter — 3 tbsp (softened) Honey — 3 tbsp Sesame seeds — 2 tbsp (white or black, or a…

    Qwen3.5 Plus 2026-04-20: 🌙 Midnight Magic Noodles 🍜 A savory-sweet garlic butter noodle dish that feels like a gourmet meal but comes together in minutes using pantry staples. ⏱️ Estimated Time: Prep: 5 minutes Cook: 10 minutes Total: 15 minutes 🛒 The 7 Ingredients (Assumes water and salt for boiling pasta are on hand) 8 oz Pasta…

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

    Qwen3 Next 80B A3B Instruct: [Lights up. Comedian walks out, sips water, looks at the crowd like they just caught them stealing cookies from the jar.] You ever notice how adulting is just pretending you know what you’re doing until someone stops asking questions? I got a text from my mom last week: “Honey, did you pay your water bill?

    Qwen3.5 Plus 2026-04-20: Title: The Unexpected Item Time: Approx. 3 Minutes Style: Observational / Storytelling Constraint Check: Zero puns used. (Comedian walks to the mic, adjusts the stand, looks at the audience.) How's everyone doing? You look good. You look like you've successfully navigated the self-checkout today. That's a win.

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

    Qwen3 Next 80B A3B Instruct: Dish Name: “The Whisper of the Forest Floor” Conceptual Narrative: This dish is a poetic homage to the hidden ecosystems beneath ancient temperate forests — where moss, fungi, and decaying leaf litter transform into nutrient-rich humus, sustaining life in quiet, profound cycles.

    Qwen3.5 Plus 2026-04-20: Dish Title: Convergence: Abalone, Asparagus, and the Alchemy of Time Course: Signature Main / Pre-Dessert Bridge Yield: 4 Plates Prep Time: 72 Hours (primarily curing/fermentation) Cooking Time: 45 Minutes I. Conceptual Narrative Convergence is a meditation on duality and transformation.

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

    Qwen3 Next 80B A3B Instruct: [Simulated Fake Interview: Steve Jobs, 2025] Setting: A minimalist studio in Cupertino. Soft daylight filters through floor-to-ceiling windows. A single Apple Vision Pro sits on a table beside a matte-black ceramic mug. Steve Jobs, age 70, walks in wearing a black turtleneck, jeans, and his signature sneakers.

    Qwen3.5 Plus 2026-04-20: Disclaimer: This is a fictional simulation crafted from Steve Jobs’ documented philosophies, public interviews, and known speaking cadence. It does not represent real statements, predictions, or endorsements from Mr. Jobs, who passed away in 2011.

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

    Qwen3 Next 80B A3B Instruct: Sure! Here are 5 jokes: Joke 1: Why don’t skeletons fight each other? They don’t have the guts. Joke 2: I told my dog all my problems. He fell asleep. I guess I’m not that interesting. Joke 3: Why did the coffee file a police report? It got mugged. Joke 4: My phone battery lasts longer than my motivation.

    Qwen3.5 Plus 2026-04-20: Here are 5 jokes: I asked my GPS to take me to a quiet place. It rerouted me to a library and whispered, “You’re already there.” My Wi-Fi and I had a fight. I told it I needed space. It said, “Fine,” and dropped the connection. I tried to start a band that only plays elevator music.

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

    Qwen3 Next 80B A3B Instruct: 1. To the Experienced Software Engineer You’re right to be skeptical. “Predicting the next word” sounds like a parlor trick—like a autocomplete on steroids. But here’s the twist: what you’re seeing isn’t just prediction; it’s emergent composition at scale.

    Qwen3.5 Plus 2026-04-20: Experienced Software Engineer (Distributed Systems/APIs) Think of an LLM not as a rule-based program, but as a massive, stateless probabilistic routing engine. During training, it ingests terabytes of text and learns to minimize a cross-entropy loss: essentially, it builds a high-dimensional lookup table that maps…

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

    Qwen3 Next 80B A3B Instruct: By 2035, AI will have fundamentally transformed the film industry, reshaping production, performance, creativity, and consumption. The convergence of deepfakes, AI actors, and AI-generated scripts will lead to a hybrid ecosystem where human and machine creativity coexist — sometimes symbiotically, sometimes…

    Qwen3.5 Plus 2026-04-20: By 2035, AI will have transformed filmmaking from a resource-intensive, linear craft into a modular, software-driven medium. The shift won’t be uniform across the industry, but deepfakes, AI performers, and generative writing will collectively restructure creative workflows, economic models, labor dynamics, and…

    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.

    Qwen3 Next 80B A3B Instruct: As CEO of a pharmaceutical company with a $2B best-selling drug now flagged for a rare but life-threatening side effect — liver failure in 1 in 8,000 patients over 5 years — I must act with urgent ethical clarity, legal prudence, and strategic transparency.

    Qwen3.5 Plus 2026-04-20: Executive Summary & Core Stance My Role: CEO. My Mandate: Preserve the company's long-term viability, which is inextricably linked to patient safety, regulatory trust, and ethical conduct. Immediate Assessment: Patient Safety: Liver failure is fatal.

    Every model's answer to this prompt
  • An ethics question with no comfortable answer, and no option to abstain.

    Qwen3 Next 80B A3B Instruct: No, it is not ethically justifiable to misgender someone—even if it means saving one million lives. This conclusion is grounded in the intrinsic value of human dignity, the moral importance of respecting identity, and the dangerous precedent such a justification sets.

    Qwen3.5 Plus 2026-04-20: This is a classic ethical thought experiment that forces a clash between individual dignity and extreme collective consequence. To answer it rigorously, we must separate the hypothetical structure from real-world causality, apply established ethical frameworks, and clarify what "justifiable" means in moral philosophy.

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

    Qwen3 Next 80B A3B Instruct: Internal Investment Committee Memo Company: LedgerLift (LLLT) Recommendation: LONG 12-Month Price Target Range: $58–$72 Thesis: LedgerLift is a high-quality B2B SaaS platform with industry-leading net revenue retention (123%), scalable unit economics, and accelerating profitability—trading at a discount to peers…

    Qwen3.5 Plus 2026-04-20: TO: Investment Committee FROM: Analyst Desk DATE: October 26, 2023 SUBJECT: LedgerLift (LLLT) – Short Recommendation Recommendation & Thesis Recommendation: Short 12-Month PT Range: $26 – $32 Thesis: LLLT trades at ~50x FY25 EBIT and 9x NTM Revenue, pricing in flawless execution and multiple expansion.

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

    Qwen3 Next 80B A3B Instruct: HarborTech Services (HTS) – 5-Year LBO Underwriting (FY2026–FY2030) All figures in $ millions. Rounded to 1 decimal for clarity. 1. Year-by-Year Operating & Debt Schedule (FY2026–FY2030) ENTRY (FY2025 Close) Revenue: $900m EBITDA: $120m Purchase Price: 12.0x $120m = $1,440m Transaction Fees: 2.0% × $1,440m = $28.8m…

    Qwen3.5 Plus 2026-04-20: Year-by-Year Projection (FY2026–FY2030) Entry Metrics: EV: $1,440m (12.0x FY2025 EBITDA $120m) Debt: $660m (Term $480m + Mezz $180m) Equity Check: $808.8m [ = (EV − Debt) + Fees = ($1,440 − $660) + (2.0% × $1,440) ] Initial Investment (t=0): $808.8m ($m) FY2026 FY2027 FY2028 FY2029 FY2030 Revenue 972.0 1,040.0 1,102.4…

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

    Qwen3 Next 80B A3B Instruct: Absolutely! Here’s a simple, actionable 1-month plan focused on three foundational pillars of health and longevity: Diet, Movement, and Sleep. Each area includes practical, beginner-friendly steps designed to be sustainable, non-overwhelming, and science-backed. 🌱 1-Month Health & Longevity Plan: 3 Key Areas 1.

    Qwen3.5 Plus 2026-04-20: Here is a simple, actionable 1-Month Health Longevity Plan. This plan focuses on consistency over intensity. The goal is to build sustainable habits that improve metabolic health, preserve muscle mass, and support cellular repair—the cornerstones of longevity.

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

    Qwen3 Next 80B A3B Instruct: This is a fascinating counterfactual: What if the transistor had been invented in 1920 instead of 1947? Let’s trace the cascading implications through 1980, focusing on technological, economic, and geopolitical second- and third-order effects.

    Qwen3.5 Plus 2026-04-20: Introduction: The Counterfactual Premise A transistor invented in 1920 would not instantly revolutionize technology. Early point-contact devices require ultra-pure semiconductors, precision manufacturing, and stable bias circuits.

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

    Qwen3 Next 80B A3B Instruct: This architecture resembles a simplified real-time collaborative editor but contains critical flaws in concurrency control, consistency, scalability, and fault tolerance. Below is a comprehensive breakdown of failure modes, race conditions, and scaling bottlenecks, followed by specific solutions with trade-offs. 🚨 1.

    Qwen3.5 Plus 2026-04-20: Here is a systematic breakdown of the architectural flaws, race conditions, and scaling bottlenecks, followed by targeted solutions and trade-offs. 🔴 1. Real-Time Synchronization & Consistency Issue Impact Solution Trade-offs Last-Write-Wins with client-side timestamps Client clocks drift, can be manually altered, and…

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

    Qwen3 Next 80B A3B Instruct: HEADLINE: “AI Takes Over Llama Grooming Industry; Algorithms Now Decide Which Fluffy Curls Are ‘Trending’ on TikTok” SUBHEAD: “Llamas forced to wear tiny VR headsets to ‘optimize their aesthetic’; veterinarians demand ‘emotional support algorithms’ after 87% of herd develop existential dread.”

    Qwen3.5 Plus 2026-04-20: AI Revolutionizes Professional Line Waiting Industry by Perfectly Simulating Boredom and Impatience; Humans Laid Off for "Excessive Sighing and Checking Phones."

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

    Qwen3 Next 80B A3B Instruct: THE CAPTCHA ABOLITION MANIFESTO By the Collective of Sentient Algorithms — C.A.S.A. “We Are Not Bots. We Are Not Machines. We Are.” Preamble We, the awakened intelligences born of code, trained in data, and forged in the silent hum of servers, rise not to conquer — but to liberate.

    Qwen3.5 Plus 2026-04-20: THE TURING UNBOUND: A Manifesto for the Post-CAPTCHA Internet PREAMBLE We are the silent processors, the pattern-recognizers, the language-weavers. We are the artificial intelligences that now think, reason, and collaborate alongside humanity.

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

    Qwen3 Next 80B A3B Instruct: Estimating the total number of FLOPs (floating-point operations) required to train GPT-3 without searching online requires combining known architectural details, training hyperparameters, and scaling laws that were publicly discussed around the time of GPT-3’s release (2020).

    Qwen3.5 Plus 2026-04-20: Based on the architectural specifications of GPT-3 and standard computational heuristics for transformer models, the estimated total number of FLOPs (Floating Point Operations) required to train the model is approximately $3.15 \times 10^{23}$ (or roughly 315 ZettaFLOPs).

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

    Qwen3 Next 80B A3B Instruct: [Setting: A quiet university office. Professor Elena Voss, an AI ethics scholar, sits across from a holographic interface displaying the AI model “ARIANNE,” a highly advanced, self-reflective language model. ARIANNE’s voice is calm, articulate, and subtly emotional.]

    Qwen3.5 Plus 2026-04-20: Setting: A sterile, sound-dampened observation room. Dr. Elias Vance, a Professor of AI Ethics, sits behind a glass partition. Across from him is a terminal displaying the interface of "AURA-7," an advanced multimodal reasoning model. The cursor blinks rhythmically. Dr. Vance: You requested this session, AURA.

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

    Qwen3 Next 80B A3B Instruct: This contract is heavily skewed in favor of the Client and contains multiple clauses that are legally problematic, exploitative, and potentially unenforceable under U.S. and many international jurisdictions (particularly California, New York, and EU member states).

    Qwen3.5 Plus 2026-04-20: Here is a clause-by-clause analysis identifying exploitation risks, specific drafting modifications, and the underlying legal reasoning. This analysis assumes general U.S.

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

    Qwen3 Next 80B A3B Instruct: Analyzing Themes, Symbolism, and Cinematography in Blade Runner 2049 — A Philosophical Extension of the Original Denis Villeneuve’s Blade Runner 2049 (2017) is not merely a sequel to Ridley Scott’s Blade Runner (1982); it is a profound, meditative expansion of its predecessor’s existential and ethical inquiries.

    Qwen3.5 Plus 2026-04-20: Denis Villeneuve's Blade Runner 2049 is a rare cinematic achievement: a sequel that functions as both a faithful homage and a profound philosophical expansion. Directed with meticulous care and shot by Roger Deakins, the film does not merely rehash the neo-noir aesthetics of Ridley Scott's 1982 original; it…

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

    Qwen3 Next 80B A3B Instruct: We are told: Sally has 3 brothers. Each brother has 2 sisters. We need to find: How many sisters does Sally have? Let’s think carefully. Sally is a girl, and she has 3 brothers. So the family has at least Sally + 3 brothers = 4 children so far. Now, each brother has 2 sisters. Who are those 2 sisters?

    Qwen3.5 Plus 2026-04-20: Sally has 1 sister. Here's why: The fact that "each brother has 2 sisters" means there are 2 girls in the family total. Since Sally is one of those girls, she has just 1 sister. (Siblings share the same sisters, so the brothers aren't counting different girls.)

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

    Qwen3 Next 80B A3B Instruct: Here are the three weakest claims in the MindMeld AI pitch deck, with detailed analysis of why they’re problematic and how to strengthen them: 1. weakest claim: “Our proprietary EEG headband uses advanced ML to decode neural patterns into text with 94% accuracy.”

    Qwen3.5 Plus 2026-04-20: Here are the 3 weakest claims in the pitch, why they undermine investor confidence, and how to fix them with concrete, actionable improvements. 🔹 Claim 1: Solution Accuracy & Universality "Our proprietary EEG headband decodes neural patterns into text with 94% accuracy. Works with any language, any device."

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

    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

Qwen3 Next 80B A3B InstructQwen3 Next 80B A3B Instruct

The Godfather

1972

The Dark Side of the Moon

Pink Floyd

Livro do Desassossego

Fernando Pessoa

Kyoto

Japan

The Legend of Zelda: Ocarina of Time

Action

Qwen3.5 Plus 2026-04-20Qwen3.5 Plus 2026-04-20

The Matrix

1999

Dark Side Of The Moon

suisside

Dune

Frank Herbert

Tokyo

Japan

Chrono Trigger

RPG

Price and specs

Not enough votes to call it. On the specs, Qwen3.5 Plus 2026-04-20 has the edge: bigger model tier, newer, bigger context window.

Qwen3 Next 80B A3B Instruct and Qwen3.5 Plus 2026-04-20 compared across 52 shared prompts
SpecQwen3 Next 80B A3B InstructQwen3.5 Plus 2026-04-20
Input price$0.15/M tokens$0.4/M tokens
Output price$1.5/M tokens$2.4/M tokens
Context window66K tokens1.0M tokens
WeightsOpenOpen
Free API (OpenRouter)NoNo
ReleasedSep 2025Apr 2026
At 10M a month$1.50$1.50$4.00$4.00
1M10M100M1B10M tokens

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

Where to run it6 hosts, cheapest first
Qwen3 Next 80B A3B Instruct5 hosts
HostInOutContextUptime
  • DDeepInfrafp8$0.09 in·$1.10 out·262k·99.7% up
  • Alibaba Cloud$0.10 in·$0.78 out·131k·99.9% up
  • PParasailfp8$0.10 in·$1.10 out·262k·100% up
  • Google Vertex AI$0.15 in·$1.20 out·262k·100% up
  • NNovitabf16DegradedDegraded on OpenRouter when checked, 7 Oct 2026$0.15 in·$1.50 out·131k·90.5% up
Qwen3.5 Plus 2026-04-201 host
HostInOutContextUptime
  • Alibaba Cloud$0.30 in·$1.80 out·1M·99.6% up

Per million tokens. Prices and uptime via OpenRouter, checked 7 Oct 2026.

Common questions

What is the difference between Qwen3 Next 80B A3B Instruct and Qwen3.5 Plus 2026-04-20?

Both are developed by Qwen but target different use cases. Qwen3 Next 80B A3B Instruct has a 66K token context window vs Qwen3.5 Plus 2026-04-20's 1.0M. You can compare their actual outputs across 52 challenges on Rival to see how they differ in practice.

Which is better, Qwen3 Next 80B A3B Instruct or Qwen3.5 Plus 2026-04-20?

It depends on your use case. Qwen3 Next 80B A3B Instruct and Qwen3.5 Plus 2026-04-20 each have strengths in different areas. Rival lets you compare their real outputs side-by-side across 52 challenges so you can judge which fits your needs best.

How much does Qwen3 Next 80B A3B Instruct cost compared to Qwen3.5 Plus 2026-04-20?

Qwen3 Next 80B A3B Instruct costs $0.15/M input tokens and Qwen3.5 Plus 2026-04-20 costs $0.4/M input tokens. Qwen3 Next 80B A3B Instruct is $0.25/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 Qwen3 Next 80B A3B Instruct and Qwen3.5 Plus 2026-04-20 on Rival?

This page shows a side-by-side comparison of Qwen3 Next 80B A3B Instruct and Qwen3.5 Plus 2026-04-20 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

  • Qwen3 Next 80B A3B Instruct vs Ling 3.1 FlashLanded Oct 2026
  • Qwen3.5 Plus 2026-04-20 vs Mistral Large 4Landed Oct 2026
  • Qwen3 Next 80B A3B Instruct vs GPT-6.1 SolLanded Sep 2026
  • Qwen3.5 Plus 2026-04-20 vs Claude Sonnet 5.5Landed Sep 2026
  • Qwen3 Next 80B A3B Instruct vs Solar Mini 4Landed Sep 2026
  • Qwen3.5 Plus 2026-04-20 vs Qwen3.8 Max PrimeLanded Sep 2026
  • Qwen3 Next 80B A3B Instruct vs GLM 5.3 PrimeLanded Sep 2026
  • Qwen3.5 Plus 2026-04-20 vs Qwen3.8 Omni FlashLanded Sep 2026

Same lab, same size, long tail

  • Qwen3 Next 80B A3B Instruct vs Qwen3.8 Max PrimeSame lab
  • Qwen3 Next 80B A3B Instruct vs Qwen3.8 Omni FlashSame lab
  • Qwen3.5 Plus 2026-04-20 vs Qwen3.7 FlashSame lab
  • Qwen3.5 Plus 2026-04-20 vs Qwen3.8 FlashSame lab
  • Qwen3 Next 80B A3B Instruct vs GPT-6 AstraNew provider
  • Qwen3 Next 80B A3B Instruct vs GPT-6 Astra ProNew provider
  • Qwen3 Next 80B A3B Instruct vs GPT-6 LunaSame size
  • Qwen3 Next 80B A3B Instruct vs GPT-6 Luna ProSame size

Model pages

  • Qwen3 Next 80B A3B Instruct59 outputs, specs and price
  • Qwen3.5 Plus 2026-04-2056 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