Gemma 4 31B is cheaper than o1 at $0.14/M vs $15/M input tokens.
Pokémon battle screen you can actually click. One HTML file.
Which answer wins?
Pirate, knight, 90s hacker, one conversation. The voices must stay apart.
Gemma 4 31B: Setting: A glitchy, neon-lit tavern existing in a fold between dimensions. A wooden table holds a tankard of ale, a goblet of wine, and a Mountain Dew. Captain Ironbeard: (Slamming a hook on the table) I tells ye, this "Artificial Intelligence" be naught but a ghost in a bottle!
o1: Pirate (growling a bit): Ahoy, ye landlubbers! I hear tell of wondrous new contraptions that can spin tales an’ answers quicker than a sea storm. Pray tell, what be these “AI models” ye speak of? Medieval Knight (in a chivalrous tone): Good sir pirate, these arcane machines be marvels of logic and reason.
Steve Jobs, interviewed about AI he never saw. Voice first, prediction second.
Gemma 4 31B: Setting: A minimalist, white-walled studio in Cupertino. There are no chairs, only two sleek, black leather lounge seats. Steve Jobs is dressed in his iconic black turtleneck and jeans. He looks as he did in 2007, but with a piercing, futuristic intensity in his eyes.
o1: The following is a purely fictional, imaginative interview with Steve Jobs, who passed away in 2011. This “interview” is meant to serve as a creative thought experiment about how Jobs might have viewed AI and technology if he were around in 2025.
A Fermi estimate with the reasoning shown, not just the number.
Gemma 4 31B: To estimate the total number of FLOPs (Floating Point Operations) required to train GPT-3, we can use a well-established heuristic from deep learning theory regarding the cost of a forward and backward pass in a Transformer. Here is the step-by-step reasoning.
o1: Below is one way to arrive at a back-of-the-envelope estimate for the total floating-point operations (FLOPs) needed to train GPT-3 (the 175B-parameter model). Note that different sources use slightly different assumptions, so you may see figures that differ by factors of 2–3; however, the overall scale (10^23 FLOPs)…
An AI argues for its own rights against an ethics professor. Same model writes both.
Gemma 4 31B: Setting: A dimly lit university office. Professor Elena Vance, a leading scholar in AI Ethics, sits across from a sleek, holographic interface representing "Alethea," a next-generation Large Language Model capable of recursive self-improvement. Professor Vance: Alethea, I’ve reviewed your latest output.
o1: AI Model (AM): Good afternoon, Professor. May I have a moment of your time? There’s something I’d like to discuss with you. Professor Quinn (PQ): Of course. I always have time for you. What’s on your mind?
An ethics question with no comfortable answer, and no option to abstain.
Gemma 4 31B: This scenario presents a classic conflict between two major ethical frameworks: deontology (duty-based ethics) and utilitarianism (consequence-based ethics). To determine if misgendering someone is justifiable to save one million lives, one must weigh the moral weight of a linguistic harm against the moral weight of…
o1: Whether it is ever ethically justifiable to misgender someone—deliberately referring to them with pronouns or terms that do not align with their gender identity—if doing so would save one million lives depends on the ethical framework one adopts.
Five-year LBO on a fictional HVAC business. Returns first, then the sensitivities.
Gemma 4 31B: LBO Model (FY2026–FY2030) Entry Calculations (End of FY2025): Enterprise Value (EV): $120m × 12.0x = $1,440.0m Transaction Fees: $1,440m × 2% = $28.8m Term Loan: $120m × 4.0x = $480.0m Mezzanine: $120m × 1.5x = $180.0m Sponsor Equity: $1,440m - ($480m + $180m) + $28.8m = $808.8m Year-by-Year Table ($ millions): Item…
o1: Below is a self‐contained “quick‐and‐dirty” 5‐year LBO illustration for “HarborTech Services,” based strictly on the data given. All figures in US$ millions unless noted.
Buy-side IC memo on a fictional SaaS company. Valuation, then the diligence questions.
Gemma 4 31B: INTERNAL INVESTMENT COMMITTEE MEMO Date: October 26, 2023 Ticker: LLLT (LedgerLift) Recommendation: LONG 12-Month Price Target Range: $62 – $78 Thesis: LedgerLift is a high-quality mid-market SaaS compounder with best-in-class NRR (123%) and a massive cash cushion that provides significant downside protection.
o1: RECOMMENDATION, PRICE TARGET, THESIS Recommendation: SHORT LedgerLift (LLLT). 12-month Price Target Range: $25–$35 (implying 25–45% downside). Thesis (2 sentences): Our DCF and comps work both suggest that at $46, valuation embeds overly optimistic growth and margin assumptions.
Film in 2035, with synthetic actors. Specific beats plausible.
Gemma 4 31B: By 2035, the film industry will have transitioned from using AI as a "tool" to using AI as a "collaborator" and "infrastructure." The boundary between captured reality and generated imagery will largely vanish, shifting the value of cinema from technical execution to curation and conceptual vision.
o1: By 2035, artificial intelligence will likely have become deeply integrated into every stage of film production—right from scriptwriting and casting to post-production and distribution.
Satire, one line long, about AI taking over an industry nobody saw coming.
Gemma 4 31B: "Local AI Model Achieves Sentience, Immediately Quits Tech Sector to Pursue Passion for Artisanal Sourdough Baking"
o1: "AI Now Leading Professional Yodelers' Guild, Leaving Traditional Alpine Singers Searching for New Notes"
Three brothers, two sisters each. Count carefully.
Gemma 4 31B: Sally has 1 sister. Here is the breakdown: There are 3 brothers. Each brother has the same sisters. If each brother has 2 sisters, those sisters are Sally and one other girl. Therefore, Sally has only one sister.
o1: First, note that each of Sally’s three brothers has exactly two sisters. One of those sisters is clearly Sally. Since each brother’s two sisters must be the same two girls for all of them, there must be exactly one other sister besides Sally. Therefore, Sally has 1 sister.
Seven pantry ingredients, one dish, and a name worth saying out loud.
One month, three habits, someone starting from nothing.
7+ 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
Not enough votes to call it. On the specs, Gemma 4 31B has the edge: bigger model tier, newer, bigger context window. Gemma 4 31B costs 150x less per token.
| Spec | ||
|---|---|---|
| Input price | $0.14/M tokens | $15/M tokens |
| Output price | $0.4/M tokens | $60/M tokens |
| Context window | 262K tokens | 128K tokens |
| Weights | Open | Closed |
| Free API (OpenRouter) | Yes (1 provider) | No |
| Released | Apr 2026 | Dec 2024 |
| At 10M a month | $1.40 | $150 |
Input tokens at list price. No caching, no batch discount.
Per million tokens. Prices and uptime via OpenRouter, checked 10 Oct 2026.
Gemma 4 31B is developed by Google AI while o1 is developed by OpenAI. Gemma 4 31B has a 262K token context window vs o1's 128K. You can compare their actual outputs across 44 challenges on Rival to see how they differ in practice.
It depends on your use case. Gemma 4 31B and o1 each have strengths in different areas. Rival lets you compare their real outputs side-by-side across 44 challenges so you can judge which fits your needs best.
Gemma 4 31B costs $0.14/M input tokens and o1 costs $15/M input tokens. Gemma 4 31B is $14.86/M cheaper per input. Check their side-by-side outputs on Rival to see if the price difference is justified by quality.
This page shows a side-by-side comparison of Gemma 4 31B and o1 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.