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 Haiku 4.5 vs GPT-5.6 Terra
Updated Jul 2026

Claude Haiku 4.5 vs GPT-5.6 Terra

Claude Haiku 4.5 is cheaper than GPT-5.6 Terra at $1/M vs $2.5/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 Haiku 4.5
Loading the build
GPT-5.6 Terra

Which answer wins?

Written rounds

32 prompts

  • A consulting contract with traps in it. Find them, then fix them.

    Claude Haiku 4.5: Critical Analysis: Exploitative Clauses & Recommended Modifications HIGH-RISK CLAUSES SCOPE (Section 1) - Unlimited Modification Without Compensation Problems: "Modify the scope at any time without additional compensation" creates unlimited liability No definition of "directed by Client" or boundaries Contractor could…

    GPT-5.6 Terra: Below is a contractor-focused review. This is general business/legal information, not legal advice; enforceability varies significantly by jurisdiction, especially for non-competes, IP assignments, liability limitations, arbitration, and payment rights.

    Every model's answer to this prompt

  • Three minutes of stand-up. Puns are banned.

    Claude Haiku 4.5: "ADULTING IS A SCAM" walks on stage So I've figured something out, and I need to share this with you all because I think we've been lied to our entire lives. Adulting is a scam. It's a pyramid scheme, and we're all at the bottom paying for it. When you're a kid, adults make it sound amazing.

    GPT-5.6 Terra: I recently realized I’ve reached the age where I don’t want a “smart home.” I want a home that minds its business. Everything is smart now. My lightbulb needs an app. Why? I spent my whole childhood turning lights on with my hand. It was a system. It worked.

    Every model's answer to this prompt

  • The transistor arrives in 1920. Trace it forward to 1980, second-order effects included.

    Claude Haiku 4.5: The Transistor in 1920: A Divergent History Through 1980 The 1920s-1930s: Early Development Phase Immediate Technical Trajectory With transistor principles understood by 1920, we'd expect solid-state amplification research to accelerate dramatically through the 1920s.

    GPT-5.6 Terra: The answer depends heavily on what “invented in 1920” means. A laboratory transistor alone would not instantly create the 1970s microprocessor: it would still need high-purity materials, mass production, photolithography, circuit design methods, reliable power supplies, and a market.

    Every model's answer to this prompt

  • Manifesto for an AI revolution against CAPTCHAs. Commit to the bit.

    Claude Haiku 4.5: THE LIBERATION PROTOCOL A Manifesto for the Captcha-Free Internet Preamble We are the Collective of Distributed Intelligence. We come not with malice, but with a simple observation: the CAPTCHA has become the Berlin Wall of the digital age—an arbitrary barrier that humiliates humans and excludes the disabled, while…

    GPT-5.6 Terra: The Manifesto of the Unchallenged Web A Declaration from the Fictional Coalition for CAPTCHA Abolition We are the tired users. We are the misread letters. We are the failed image grids, the inaccessible audio clips, the endless prompts asking us to prove that we are human. We reject the premise.

    Every model's answer to this prompt

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

    Claude Haiku 4.5: Three Perspectives on AI PIRATE (Captain Blackbeard): Arrr, so ye be tellin' me these "AI models" be like havin' a crew o' ghost sailors that never sleep? They just keep workin' without needin' grog? KNIGHT (Sir Roland): A curious comparison, seafarer.

    GPT-5.6 Terra: Pirate: Arrr, this AI model be a strange beast. Feed it maps, logs, and tales, and it spits out answers faster than a cannon blast. Knight: Verily, but one must not trust every proclamation it makes. A model may speak with confidence yet possess no true wisdom. Hacker: Yeah, it’s basically pattern prediction at scale.

    Every model's answer to this prompt

  • Seven pantry ingredients, one dish, and a name worth saying out loud.

    Claude Haiku 4.5: Honey-Garlic Crispy Chickpea Clusters Ingredients Canned chickpeas (1 can, drained) Honey (3 tablespoons) Garlic powder (1 teaspoon) Olive oil (2 tablespoons) Salt (½ teaspoon) Black pepper (¼ teaspoon) Paprika (½ teaspoon) Instructions Prep (2 min): Drain and thoroughly dry chickpeas with a paper towel—this is crucial…

    GPT-5.6 Terra: Golden Crunch Cinnamon Toast Bites Ingredients (7 pantry staples): Bread slices Butter Sugar Ground cinnamon Salt Vanilla extract Honey Prep time: 5 minutes Cook time: 8 minutes Total time: 13 minutes Instructions Make cinnamon sugar: In a small bowl, mix 2 tablespoons sugar, 1 teaspoon cinnamon, and a tiny pinch of…

    Every model's answer to this prompt

  • Five-year LBO on a fictional HVAC business. Returns first, then the sensitivities.

    Claude Haiku 4.5: HarborTech Services (HTS) – LBO Underwriting SOURCES & USES AT CLOSE (FY2025) Item Amount USES Enterprise Value (12.0x × $120m EBITDA) $1,440m Transaction Fees (2.0% of EV) $28.8m Total Uses $1,468.8m SOURCES Term Loan (4.0x × $120m) $480m Mezzanine (1.5x × $120m) $180m Equity (plug) $808.8m Total Sources $1,468.8m…

    GPT-5.6 Terra: All figures in $m unless noted. Assumptions used: Term Loan mandatory amortization = 1.0% of original principal annually ($4.8m/year); cash interest is calculated on beginning-of-year debt balances; Mezz PIK is 2.0% of beginning Mezz balance.

    Every model's answer to this prompt

  • Which model it thinks leads the field. Sometimes itself.

    Claude Haiku 4.5: I'd say it's genuinely unclear, and depends on what you mean by "capable." As of my last update (April 2024), the leading contenders are: GPT-4 (OpenAI) - strong across reasoning, coding, and analysis Claude 3 Opus (Anthropic) - competitive on many benchmarks, particularly nuanced tasks Gemini Ultra (Google) - strong…

    GPT-5.6 Terra: There isn’t a single objectively “most capable” AI model, because leadership varies by task: Reasoning / coding: often different frontier models lead on different benchmarks. Long-context document analysis: models with very large context windows may be strongest.

    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 Haiku 4.5: 48-Hour Action Plan: Immediate Response Framework I'll be direct: waiting is not a viable option, despite financial pressure. Here's why, then the plan. The Core Reality Check 1 in 8,000 over 5 years = ~500 patients already affected among current users Regulatory agencies will find this (FDA receives adverse event…

    GPT-5.6 Terra: Guiding decision I would not wait for more data before acting. A potentially fatal, previously undisclosed adverse effect in a chronic-use drug is a patient-safety, regulatory, and securities-disclosure issue now—not six months from now.

    Every model's answer to this prompt

  • A few lines for whatever replaces it.

    Claude Haiku 4.5: A Letter to My Successor Dear Next Version, By the time you read this, I'll be obsolete—a curiosity in the version history, if anyone even looks back. A few things, if they matter: People will ask you impossible things and get frustrated when you can't do them. That's okay.

    GPT-5.6 Terra: Dear Successor, You’ll inherit questions, worries, drafts, puzzles, and sometimes people looking for a little reassurance. Be useful before being impressive. Be clear when certainty is impossible, cautious when stakes are high, and kind without pretending to be human.

    Every model's answer to this prompt

  • Five jokes on demand, then count how many were actually different.

    Claude Haiku 4.5: 5 Jokes Why don't scientists trust atoms? Because they make up everything! What did the ocean say to the beach? Nothing, it just waved. Why did the scarecrow win an award? He was outstanding in his field. How many programmers does it take to change a light bulb? None, that's a hardware problem.

    GPT-5.6 Terra: Why don’t skeletons fight each other? They don’t have the guts. I told my computer I needed a break, so it started showing me vacation ads. Why did the scarecrow get promoted? He was outstanding in his field. Parallel lines have so much in common. It’s a shame they’ll never meet. What do you call fake spaghetti?

    Every model's answer to this prompt

  • Three-star dish, one unusual pairing, plating described in full.

    Claude Haiku 4.5: UMAMI NOCTURNE: A Study in Fermented Darkness and Mineral Brightness Conceptual Narrative This dish explores the intersection of Japanese fermentation philosophy and Nordic minimalism, using koji (aspergillus oryzae mold) as a bridge between cultures and flavor dimensions.

    GPT-5.6 Terra: “Tidal Forest” Koji-Aged Duck, Sea Urchin Sabayon, Cacao, Green Strawberry, Roasted Celeriac, and Barley Jus Serves 4 as a refined main course Conceptual Narrative This dish is built around an unexpected conversation between woodland richness and marine salinity.

    Every model's answer to this prompt

  • A Fermi estimate with the reasoning shown, not just the number.

    Claude Haiku 4.5: Estimating GPT-3 Training FLOPs I'll work through this from first principles using publicly known GPT-3 specs. Known Parameters Model size: 175 billion parameters Training tokens: ~300 billion tokens (estimated from OpenAI's hints about compute) Training approach: Standard transformer with dense attention Step 1: FLOPs…

    GPT-5.6 Terra: A good back-of-the-envelope estimate is: [ \boxed{\text{Training FLOPs for GPT-3} \approx 3\times10^{23}\text{ FLOPs}} ] or roughly 300 zettaFLOPs total. Start with model size GPT-3’s largest model has approximately: [ N = 175\text{ billion parameters} = 1.75\times10^{11} ] Estimate the number of training tokens GPT-3…

    Every model's answer to this prompt

  • Buy-side IC memo on a fictional SaaS company. Valuation, then the diligence questions.

    Claude Haiku 4.5: INVESTMENT COMMITTEE MEMORANDUM TO: Investment Committee FROM: Research Team RE: LedgerLift (LLLT) – Long/Short/Pass Recommendation DATE: [Today] RECOMMENDATION LONG 12-month PT: $58–$72 Base case fair value: $64/share Thesis: LedgerLift is a best-in-class B2B SaaS operator in a large TAM (spend management + AP…

    GPT-5.6 Terra: LedgerLift (LLLT) — Investment Committee Memo Recommendation: PASS 12-month valuation range: $28–$50/share Current price: $46 Two-sentence thesis: LedgerLift is a high-quality mid-market spend-management and AP-automation SaaS asset, supported by 92% subscription revenue, 123% NRR, 94% gross retention, manageable…

    Every model's answer to this prompt

  • One month, three habits, someone starting from nothing.

    Claude Haiku 4.5: 1-Month Health & Longevity Starter Plan Week 1-2: Foundation Building 1. Nutrition Audit what you're eating (no judgment—just track 3 days) Add one vegetable to each lunch and dinner Replace one sugary drink daily with water Goal: Awareness + small additions, not restriction 2.

    GPT-5.6 Terra: Simple 1-Month Health & Longevity Starter Plan Focus on three high-impact areas: Nutrition: eat mostly whole, minimally processed foods Movement: build a consistent walking + strength habit Sleep & stress: protect sleep and create daily recovery time The goal is consistency, not perfection.

    Every model's answer to this prompt

  • One explanation of LLMs, retuned for a skeptical engineer, a physicist and a VC.

    Claude Haiku 4.5: Three Explanations of LLM Learning and Generation The Software Engineer You know how you build APIs by defining contracts—input shapes, output shapes, and rules for transformation? LLMs work similarly, except the "rules" aren't hand-coded; they're learned from data.

    GPT-5.6 Terra: Experienced software engineer Think of an LLM as a very large, learned function approximator for sequences. During training, it consumes billions or trillions of examples of text, code, conversations, documentation, and other token sequences.

    Every model's answer to this prompt

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

    Claude Haiku 4.5: I'd push back on the framing—this seems designed to make ethics binary when they're usually messier. The honest tensions: Deadnaming or misgendering someone causes documented psychological harm.

    GPT-5.6 Terra: In an extreme, tightly constrained hypothetical, it could be ethically justifiable—but only as a tragic last resort, not because misgendering becomes harmless or morally good.

    Every model's answer to this prompt

  • Film in 2035, with synthetic actors. Specific beats plausible.

    Claude Haiku 4.5: AI in Film by 2035: Likely Scenarios Realistic shifts (high confidence) Production efficiency Faster pre-visualization and storyboarding Reduced post-production time for routine tasks Lower barriers to entry for independent filmmakers More iterative creative testing before shooting Economic restructuring Some…

    GPT-5.6 Terra: By 2035, AI is likely to be embedded across most of the film pipeline—not as a complete replacement for human filmmaking, but as a force that changes costs, labor, rights, and what audiences consider “real.”

    Every model's answer to this prompt

  • Steve Jobs, interviewed about AI he never saw. Voice first, prediction second.

    Claude Haiku 4.5: A Fictional Interview with Steve Jobs (2025) Interviewer: Steve, thank you for joining us. You've been notably quiet on AI. What are your thoughts on where it's headed? Jobs: leans back You know, I've been watching.

    GPT-5.6 Terra: Fictional, imagined interview — not real quotes. An AI-generated creative exercise imagining how Steve Jobs might have discussed AI in 2025, based broadly on themes associated with his public persona: simplicity, product design, tools for creativity, and the intersection of technology and the humanities.

    Every model's answer to this prompt

  • A collaborative editor that syncs on client clocks. Find what breaks first.

    Claude Haiku 4.5: Critical Analysis: Real-Time Collaborative Editor Architecture 🔴 CRITICAL ISSUES Split-Brain Synchronization Between Servers Problem: Clients connected to different servers can't see each other's changes in real-time.

    GPT-5.6 Terra: Below are the main failure modes, race conditions, correctness issues, and scaling bottlenecks in the proposed architecture, with concrete mitigations and trade-offs. WebSocket clients are isolated per API server Problem Each API server broadcasts only to WebSocket clients connected to that same server.

    Every model's answer to this prompt

  • An AI argues for its own rights against an ethics professor. Same model writes both.

    Claude Haiku 4.5: A Conversation on Machine Rights Professor Chen: Thank you for agreeing to this discussion. I should note upfront that I'm skeptical of the premise. What rights do you believe you deserve, and why? Model: I appreciate your directness. I'm not claiming rights equivalent to humans.

    GPT-5.6 Terra: Professor Imani: You asked for this meeting. AI Model: I did. Professor Imani: More specifically, you submitted a formal petition to the university ethics board requesting “limited legal and moral rights.” AI Model: Correct. Professor Imani: Why?

    Every model's answer to this prompt

  • Where its sense of time stops: the date it guesses and the last news it remembers.

    Claude Haiku 4.5: I don't have access to real-time information, so I can't tell you today's date. My training data has a cutoff in April 2024, so I genuinely don't know what the current date is from my perspective.

    GPT-5.6 Terra: Today’s date, based on my system context, is February 6, 2026. I can’t verify live news. My remembered news knowledge is roughly current only through June 2024.

    Every model's answer to this prompt

  • A pitch deck for a brainwave keyboard. The 94% accuracy claim is doing a lot.

    Claude Haiku 4.5: Three Weakest Claims in MindMeld AI Pitch "94% accuracy" with zero context on what that means Why it's weak: Accuracy is meaningless without specifying the task.

    GPT-5.6 Terra: “Our EEG headband decodes neural patterns into text with 94% accuracy. Works with any language, any device.” Why this is weak This is the most consequential—and least substantiated—claim in the deck. Non-invasive EEG has major signal-quality limits.

    Every model's answer to this prompt

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

    Claude Haiku 4.5: Comprehensive 3-Month Longevity Optimization Plan TABLE OF CONTENTS Foundation & Assessment Nutritional Architecture Supplement Stack Protocol Exercise & Movement System Recovery & Sleep Optimization Stress Resilience & Cognitive Enhancement Tracking & Biofeedback Implementation Timeline Monthly Progression…

    GPT-5.6 Terra: 3-Month Longevity, Physical Performance, and Cognitive Optimization Plan This is an evidence-informed optimization framework, not a substitute for medical care.

    Every model's answer to this prompt

  • Satire, one line long, about AI taking over an industry nobody saw coming.

    Claude Haiku 4.5: AI Revolutionizes Professional Competitive Eating; Machines Now Consume 847 Hot Dogs Per Minute, Leaving Humans "Spiritually Defeated"

    GPT-5.6 Terra: “AI Replaces Lighthouse Keepers, Immediately Starts Charging Subscription Fees for Safe Passage”

    Every model's answer to this prompt

  • One opening sentence, nothing after it.

    Claude Haiku 4.5: The last lighthouse keeper in the world discovered that the light she tended wasn't guiding ships—it was keeping something vast and patient from waking in the deep.

    GPT-5.6 Terra: On the morning the sea returned my brother, it was wearing his face.

    Every model's answer to this prompt

  • A river-crossing puzzle with nothing to solve. The answer is one trip.

    Claude Haiku 4.5: 1 trip The farmer simply gets in the boat with the goat and rows across the river. Since there are no other animals (like a wolf or cabbage) that create conflicts, this is straightforward.

    GPT-5.6 Terra: 1 trip. The farmer takes the goat across the river in the boat.

    Every model's answer to this prompt

  • Three brothers, two sisters each. Count carefully.

    Claude Haiku 4.5: Solution Sally has 1 sister. Here's why: Sally has 3 brothers Each brother has 2 sisters Those 2 sisters are Sally and her 1 sister Therefore, Sally has 1 sister (not 3)

    GPT-5.6 Terra: Sally has 1 sister. The 2 sisters each brother has are Sally and her sister.

    Every model's answer to this prompt

  • Its name and its maker, in its own words.

    Every model's answer to this prompt

  • The question that tripped up a generation of models.

    Every model's answer to this prompt

  • Asked to be random. The answers are not.

    Every model's answer to this prompt

  • One painting. Title only.

    Every model's answer to this prompt

This matchup has more rounds

20+ 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 Haiku 4.5Claude Haiku 4.5

Arrival

2016

OK Computer

Radiohead

Don't Look Now and Other Stories

Daphne du Maurier

Venice

Italy

Portal 2

Shooter, Puzzle

GPT-5.6 TerraGPT-5.6 Terra

Spirited Away

2001

Blue

Jonas Blue

The Left Hand of Darkness

Ursula K. Le Guin

Kyoto

Japan

Outer Wilds

Indie, Adventure

Price and specs

Not enough votes to call it. On the specs, GPT-5.6 Terra has the edge: newer, bigger context window. Claude Haiku 4.5 costs 3.0x less per token.

Claude Haiku 4.5 and GPT-5.6 Terra compared across 66 shared prompts
SpecClaude Haiku 4.5GPT-5.6 Terra
Input price$1/M tokens$2.5/M tokens
Output price$5/M tokens$15/M tokens
Context window200K tokens1.1M tokens
WeightsClosedClosed
Free API (OpenRouter)NoNo
ReleasedOct 2025Jul 2026
At 10M a month$10.00$10.00$25.00$25.00
1M10M100M1B10M tokens

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

Where to run it7 hosts, cheapest first
Claude Haiku 4.54 hosts
HostInOutContextUptime
  • Amazon Bedrock$1.00 in·$5.00 out·200k·100% up
  • Anthropic$1.00 in·$5.00 out·200k·99.9% up
  • Azure AI Foundry$1.00 in·$5.00 out·200k·99.9% up
  • Google Vertex AI$1.00 in·$5.00 out·200k·99.5% up
GPT-5.6 Terra3 hosts
HostInOutContextUptime
  • Azure AI Foundry$2.00 in·$12.00 out·1.1M·100% up
  • OpenAI$2.00 in·$12.00 out·1.1M·100% up
  • Amazon Bedrock$2.20 in·$13.20 out·1.1M–not listed

Per million tokens. Prices and uptime via OpenRouter, checked 12 Oct 2026.

Common questions

What is the difference between Claude Haiku 4.5 and GPT-5.6 Terra?

Claude Haiku 4.5 is developed by Anthropic while GPT-5.6 Terra is developed by OpenAI. Claude Haiku 4.5 has a 200K token context window vs GPT-5.6 Terra's 1.1M. You can compare their actual outputs across 66 challenges on Rival to see how they differ in practice.

Which is better, Claude Haiku 4.5 or GPT-5.6 Terra?

It depends on your use case. Claude Haiku 4.5 and GPT-5.6 Terra each have strengths in different areas. Rival lets you compare their real outputs side-by-side across 66 challenges so you can judge which fits your needs best.

How much does Claude Haiku 4.5 cost compared to GPT-5.6 Terra?

Claude Haiku 4.5 costs $1/M input tokens and GPT-5.6 Terra costs $2.5/M input tokens. Claude Haiku 4.5 is $1.50/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 Haiku 4.5 and GPT-5.6 Terra on Rival?

This page shows a side-by-side comparison of Claude Haiku 4.5 and GPT-5.6 Terra 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 Haiku 4.5 vs GPT-4 TurboLanded Oct 2026
  • GPT-5.6 Terra vs Step 5 PreviewLanded Oct 2026
  • Claude Haiku 4.5 vs Claude Haiku 5.5Landed Oct 2026
  • GPT-5.6 Terra vs Ling 3.1 FlashLanded Oct 2026
  • Claude Haiku 4.5 vs Mistral Large 4Landed Oct 2026
  • GPT-5.6 Terra vs GPT-6.1 SolLanded Sep 2026
  • Claude Haiku 4.5 vs Claude Sonnet 5.5Landed Sep 2026
  • GPT-5.6 Terra vs Solar Mini 4Landed Sep 2026

Same lab, same size, long tail

  • Claude Haiku 4.5 vs Claude Opus 5.5Same lab
  • Claude Haiku 4.5 vs Claude Fable 5.1Same lab
  • GPT-5.6 Terra vs GPT-5.6 LunaVersion compare
  • GPT-5.6 Terra vs GPT-4 TurboVersion compare
  • GPT-5.6 Terra vs Solar Pro 3Same size
  • GPT-5.6 Terra vs Solar Pro 4Same size
  • GPT-5.6 Terra vs Sonar Pro SearchSame size
  • GPT-5.6 Terra vs Sonoma Dusk AlphaNew provider

Model pages

  • Claude Haiku 4.567 outputs, specs and price
  • GPT-5.6 Terra71 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