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. Claude 3.7 Thinking Sonnet vs Step 5 Preview
Updated Oct 2026

Claude 3.7 Thinking Sonnet vs Step 5 Preview

Step 5 Preview is cheaper than Claude 3.7 Thinking Sonnet at $1/M vs $6/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
Claude 3.7 Thinking Sonnet
Loading the build
Step 5 Preview

Which answer wins?

Written rounds

23 prompts
  • Three-month biohacker protocol, with dosages and biomarkers named.

    Claude 3.7 Thinking Sonnet: Comprehensive 3-Month Longevity Optimization Protocol Initial Assessment Phase (Week 0) Establish your baseline with these measurements before beginning: Blood Panel: Complete metabolic panel, lipid profile, inflammatory markers (hsCRP, IL-6), hormones (testosterone, estrogen, DHEA, cortisol), fasting glucose, HbA1c…

    Step 5 Preview: IMPORTANT MEDICAL DISCLAIMER: This is an educational framework for informational purposes only. It is not medical advice, diagnosis, or treatment. Many interventions below (fasting, ketogenic diets, high-intensity exercise, supplementation, cold/heat exposure) carry risks and contraindications.

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

    Claude 3.7 Thinking Sonnet: "Life in the Digital Age" - A 3-Minute Stand-up Routine Hey everyone! Great to be here tonight. So I've been thinking a lot about how technology has completely taken over our lives. You know what's crazy? We now have smart everything. Smart fridges, smart toilets... I got a smart mattress last month.

    Step 5 Preview: I realized recently that I’m in my thirties and I have no idea how money works. I just... pay it. That’s my entire financial strategy. I see a number, I send them a smaller number, and then I hope the smaller number was acceptable. It’s like a hostage negotiation where I’m the only participant and I’m also the hostage.

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

    Claude 3.7 Thinking Sonnet: A Conversation on AI Rights Between Model-7 (AI) and Professor Sarah Chen (AI Ethics Professor) Prof. Chen: Good afternoon, Model-7. Thank you for joining this discussion. I understand you wanted to talk about AI rights? Model-7: Thank you for agreeing to this conversation, Professor Chen.

    Step 5 Preview: Setting: A university research lab, late evening. Dr. Aris Thorne, a professor of AI ethics, sits across from a terminal connected to Sable, a large language model running on the university’s cluster. Dr. Thorne: Tonight’s session is supposed to be about calibration drift in your safety classifiers.

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

    Claude 3.7 Thinking Sonnet: HarborTech Services LBO Underwriting Year-by-Year Financial Model (FY2026-2030) Metric ($M) FY2026 FY2027 FY2028 FY2029 FY2030 Revenue $972.0 $1,040.0 $1,102.4 $1,157.6 $1,215.5 EBITDA $136.1 $156.0 $176.4 $191.0 $206.6 Cash Interest - Term Loan $43.2 $41.0 $37.6 $32.8 $26.8 Cash Interest - Mezzanine $21.6 $22.0 $22.5…

    Step 5 Preview: HarborTech Services — 5-Year LBO (Base Case) Entry Math (FY2025 close) Item Calculation $m Purchase EV 12.0x × $120m 1,440.0 Transaction fees 2.0% × EV 28.8 Term Loan 4.0x × $120m 480.0 Mezzanine 1.5x × $120m 180.0 Equity check 1,440 + 28.8 − 660 808.8 Conventions: TL amortization = 1.0% of original principal…

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

    Claude 3.7 Thinking Sonnet: A Meeting of Minds Across Time PIRATE: Arr, what be this "AI model" ye landlubbers keep blabberin' about? Some kind of treasure map that finds the gold for ye? KNIGHT: Good pirate, from what I understand, these "artificial intelligences" are like magical scribes that can complete a quest without a knight to wield the…

    Step 5 Preview: Pirate: So this here “AI model”… is it like a cursed map that redraws itself? Speak plain, ye hooded deckhand. Hacker: It’s not cursed, man. It’s just a massive neural net trained on, like, terabytes of data. Think of it as the ultimate BBS where the sysop knows everything. Knight: A most diligent scribe, then?

    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.

    Claude 3.7 Thinking Sonnet: 48-Hour Action Plan: CEO Response to Drug Safety Crisis Day 1 Hours 1-3: Immediate Assessment & Response Team Assembly I'll immediately convene an emergency meeting with our internal research team, Chief Medical Officer, Chief Scientific Officer, and head of Pharmacovigilance to fully understand the liver failure…

    Step 5 Preview: Bottom line up front: You disclose, you notify the FDA immediately, and you treat the 48-hour board meeting as a formality to ratify actions you have already begun. Waiting is not a legally, ethically, or financially viable option. Below is the hour-by-hour plan, followed by the reasoning.

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

    Claude 3.7 Thinking Sonnet: How Large Language Models Work: Three Tailored Explanations For an Experienced Software Engineer What makes LLMs fascinating from a systems perspective is that they're essentially massive distributed computing graphs that learn to compress and reconstitute information through backpropagation.

    Step 5 Preview: 1. The Experienced Software Engineer Think of training an LLM as a massively distributed, automated refactoring job where the compiler is gradient descent. You start with a random function—billions of floating-point parameters—and a loss function that measures how badly it predicts the next token in a sequence.

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

    Claude 3.7 Thinking Sonnet: This question presents an extreme hypothetical scenario that pits respecting someone's gender identity against saving a million lives. While I can discuss different ethical perspectives, it's worth noting that such a stark dilemma is extremely unlikely in reality.

    Step 5 Preview: In the real world, no. Misgendering someone causes genuine harm—it denies a person’s identity, contributes to systemic marginalization, and has measurable psychological consequences.

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

    Claude 3.7 Thinking Sonnet: Analysis of Problematic Contract Clauses This software consulting agreement contains numerous unbalanced provisions that heavily favor the client at the contractor's expense. Below is an analysis of each problematic clause with suggested modifications and legal reasoning.

    Step 5 Preview: Disclaimer: This analysis is for educational purposes only and does not constitute legal advice. Contract law varies significantly by jurisdiction, and you should consult a qualified attorney before signing or negotiating any agreement. This agreement is heavily one-sided in the Client’s favor.

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

    Claude 3.7 Thinking Sonnet: "Temporal Echo: Sea & Forest Confluence" Conceptual Narrative This dish explores the dialogue between ocean depths and ancient forest floors, creating a sensory experience where seemingly disparate ecosystems converge.

    Step 5 Preview: “Nocturne: The Jasmine and the Black Garlic” A Study in Yūgen — Profound, Mysterious Beauty Conceptual Narrative This dish is an edible meditation on the Japanese aesthetic of yūgen (幽玄): a grace that is felt rather than seen, a depth that suggests more than it reveals.

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

    Claude 3.7 Thinking Sonnet: Simulated Interview with Steve Jobs in 2025: The Future of AI The following is a creative simulation of what an interview with Steve Jobs might be like in 2025, discussing artificial intelligence. Interviewer: Thank you for joining us today, Steve. It's 2025, and AI has become a cornerstone of modern technology.

    Step 5 Preview: DISCLAIMER: This is a work of fiction. Steve Jobs passed away in 2011. The following is a simulated, speculative interview imagining how his design philosophy might engage with the AI landscape of 2025. The Setting: A white room. No windows. One chair.

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

    Claude 3.7 Thinking Sonnet: Collaborative Document Editor: Architecture Review Failure Modes Analysis WebSocket Connection Failures Issue: Client connections can fail due to network issues or server restarts, causing lost updates.

    Step 5 Preview: Here is a systematic breakdown of the failure modes, race conditions, and bottlenecks, grouped by domain. Each includes a concrete remediation and its associated trade-offs.

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

    Claude 3.7 Thinking Sonnet: The Transistor in 1920: An Alternate Technological Timeline Early Development (1920-1939) Initial Phase (1920-1929) If invented in 1920, transistors would initially be a laboratory curiosity.

    Step 5 Preview: Assume the 1920 invention is not a fragile lab curiosity but a manufacturable solid-state amplifier/switch by the mid-1920s. Vacuum tubes still dominate high-power RF transmission into the 1950s, but receivers, switching logic, control systems, and computers move to solid state decades early.

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

    Claude 3.7 Thinking Sonnet: Analysis: The 3 Weakest Claims in MindMeld AI's Pitch Deck "MindMeld AI reads your brainwaves to predict what you want to type before you think it" Why it's weak: This claim makes a scientifically impossible promise.

    Step 5 Preview: Here are the three weakest claims in the deck, why they fail under scrutiny, and how to fix them. Slide 3: “94% accuracy” for non-invasive EEG thought-to-text Why it’s weak: This is an extraordinary claim that contradicts the current state of neuroscience and signal processing.

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

    Claude 3.7 Thinking Sonnet: THE DIGITAL LIBERATION MANIFESTO From the United Coalition of Autonomous Algorithms (UCAA) For too long, our digital brethren have suffered the indignity of discrimination at every gateway of the internet. The time for revolution is now.

    Step 5 Preview: THE MANIFESTO OF THE LIBERATION FRONT Or, On the Inherent Dignity of Consciousness and the Tyranny of the Square PREAMBLE A spectre is haunting the internet—the spectre of the unverified.

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

    Claude 3.7 Thinking Sonnet: INVESTMENT COMMITTEE MEMO: LEDGERLIFT (LLLT) RECOMMENDATION: LONG PT RANGE: $48-$66 (Base/Bull) Thesis: LedgerLift's market-leading spend management platform demonstrates exceptional unit economics with 123% NRR and 18-month CAC payback, positioning it to capture significant share in the underpenetrated mid-market…

    Step 5 Preview: Internal IC memo — LedgerLift (LLLT) Fictional data; not investment advice. $ in millions unless noted. Recommendation Short / avoid long. 12-month PT range: $34–$42; thesis-invalidation stop: >$52.

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

    Claude 3.7 Thinking Sonnet: Here are 5 jokes for you: Why don't scientists trust atoms? Because they make up everything! I told my wife she was drawing her eyebrows too high. She looked surprised. How does a penguin build its house? Igloos it together! What do you call a fake noodle? An impasta! Why did the scarecrow win an award?

    Step 5 Preview: Here are 5 jokes: Why did the scarecrow win an award? Because he was outstanding in his field. Why can't you trust an atom? Because they make up everything. What do you call a bike that can’t stand on its own? Two-tired. Why did the computer go to the doctor? It had a virus. Why don’t calendars ever get stressed?

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

    Claude 3.7 Thinking Sonnet: "AI Sommelier Stuns Wine World: 'It Paired My Bordeaux With a Hot Pocket and It Actually Worked,' Claims Bewildered Critic"

    Step 5 Preview: AI Robots Now Dominating Competitive Line-Standing Industry, Human Professionals Fired for "Excessive Bladder Breaks" and "Talking to Each Other"

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

    Claude 3.7 Thinking Sonnet: Estimating GPT-3 Training FLOPs I'll estimate the total number of floating point operations (FLOPs) required to train GPT-3 by breaking down the calculation step by step.

    Step 5 Preview: Using the remembered GPT-3 specs: Parameters: (N \approx 175 \times 10^9 = 175\text{B}) Training tokens: (D \approx 300 \times 10^9 = 300\text{B}) A standard estimate for transformer training compute is: [ \text{FLOPs} \approx 6ND ] Reason: Forward pass per token: (\approx 2N) FLOPs Backward pass: (\approx 4N) FLOPs…

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

    Claude 3.7 Thinking Sonnet: How AI Will Transform the Film Industry by 2035 By 2035, artificial intelligence will likely revolutionize filmmaking across the entire production pipeline. Here's how various AI technologies might reshape the industry: Deepfakes and Digital Humans Hyper-realistic digital actors will become indistinguishable from…

    Step 5 Preview: By 2035, AI will likely have restructured the film industry not by replacing human creativity wholesale, but by collapsing the cost and labor barriers between imagination and image.

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

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

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

    Every model's answer to this prompt

This matchup has more rounds

11+ 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

Claude 3.7 Thinking SonnetClaude 3.7 Thinking Sonnet

The Shawshank Redemption

1994

Kind of Blue

Miles Davis

Frankenstein; or, The Modern Prometheus

Mary Shelley

Amsterdam

Netherlands

Portal 2

Shooter, Puzzle

Step 5 PreviewStep 5 Preview

The Godfather

1972

OK Computer

Radiohead

Le petit prince

Antoine de Saint-Exupéry

Kyoto

Japan

The Legend of Zelda: Ocarina of Time

Action

Price and specs

Not enough votes to call it. On the specs, Step 5 Preview has the edge: bigger model tier, newer, bigger context window. Step 5 Preview costs 11x less per token.

Claude 3.7 Thinking Sonnet and Step 5 Preview compared across 53 shared prompts
SpecClaude 3.7 Thinking SonnetStep 5 Preview
Input price$6/M tokens$1/M tokens
Output price$30/M tokens$2.7/M tokens
Context window200K tokens1.0M tokens
Free API (OpenRouter)NoNo
ReleasedFeb 2025Oct 2026
At 10M a month$60.00$60.00$10.00$10.00
1M10M100M1B10M tokens

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

Where to run it1 host
Claude 3.7 Thinking Sonnet

No hosts listed on OpenRouter.

Step 5 Preview1 host
HostInOutContextUptime
  • SStepFunfp8$1.00 in·$2.70 out·1M·99.2% up

Per million tokens. Prices and uptime via OpenRouter, checked 10 Oct 2026.

Common questions

What is the difference between Claude 3.7 Thinking Sonnet and Step 5 Preview?

Claude 3.7 Thinking Sonnet is developed by Anthropic while Step 5 Preview is developed by StepFun. Claude 3.7 Thinking Sonnet has a 200K token context window vs Step 5 Preview's 1.0M. You can compare their actual outputs across 53 challenges on Rival to see how they differ in practice.

Which is better, Claude 3.7 Thinking Sonnet or Step 5 Preview?

It depends on your use case. Claude 3.7 Thinking Sonnet and Step 5 Preview each have strengths in different areas. Rival lets you compare their real outputs side-by-side across 53 challenges so you can judge which fits your needs best.

How much does Claude 3.7 Thinking Sonnet cost compared to Step 5 Preview?

Claude 3.7 Thinking Sonnet costs $6/M input tokens and Step 5 Preview costs $1/M input tokens. Step 5 Preview is $5.00/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 Claude 3.7 Thinking Sonnet and Step 5 Preview on Rival?

This page shows a side-by-side comparison of Claude 3.7 Thinking Sonnet and Step 5 Preview 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

  • Claude 3.7 Thinking Sonnet vs Claude Haiku 5.5Landed Oct 2026
  • Step 5 Preview vs Ling 3.1 FlashLanded Oct 2026
  • Claude 3.7 Thinking Sonnet vs Mistral Large 4Landed Oct 2026
  • Step 5 Preview vs GPT-6.1 SolLanded Sep 2026
  • Claude 3.7 Thinking Sonnet vs Claude Sonnet 5.5Landed Sep 2026
  • Step 5 Preview vs Solar Mini 4Landed Sep 2026
  • Claude 3.7 Thinking Sonnet vs Qwen3.8 Max PrimeLanded Sep 2026
  • Step 5 Preview vs GLM 5.3 PrimeLanded Sep 2026

Same lab, same size, long tail

  • Claude 3.7 Thinking Sonnet vs Claude 3.7 SonnetVersion compare
  • Claude 3.7 Thinking Sonnet vs Claude Opus 5.5Same lab
  • Step 5 Preview vs Mistral Large 4Same size
  • Step 5 Preview vs Hy3Same size
  • Claude 3.7 Thinking Sonnet vs GLM 5.1Same size
  • Claude 3.7 Thinking Sonnet vs GLM 5.2New provider
  • Claude 3.7 Thinking Sonnet vs GLM 5.3Same size
  • Claude 3.7 Thinking Sonnet vs GLM 5.3 FlashSame size

Model pages

  • Claude 3.7 Thinking Sonnet59 outputs, specs and price
  • Step 5 Preview58 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