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. /
  3. Compare
  4. /
  5. GPT-5.4 MinivsMercury 2.5
Updated Sep 2026

GPT-5.4 MinivsMercury 2.5

Mercury 2.5 is cheaper than GPT-5.4 Mini at $0.04/M vs $0.75/M input tokens.

GPT-5.4 Mini and Mercury 2.5 compared across 20 shared prompts
SpecGPT-5.4 MiniMercury 2.5
Input price$0.75/M tokens$0.04/M tokens
Output price$4.5/M tokens$0.15/M tokens
Context window400K tokens260K tokens
WeightsClosed—
Free API (OpenRouter)NoNo
ReleasedMar 2026Sep 2026
Side by side

Same prompt, both answers

Opening lines from 6 of the 20 prompts both answered. Formatting stripped, words untouched.

Explain Like I'm a Specific Expert

Explain how a large language model (like GPT or Claude) learns and generates text to each of these three audiences: 1. **An experienced…

GPT-5.4 Mini893 words
For an experienced software engineer A large language model is trained to do one very specific thing: given a sequence of tokens, estimate the probability distribution of the next token. “Token” means a chunk of text, often a word piece rather than a full word.
Read the full answer
Mercury 2.5572 words
For the Experienced Software Engineer Think of the model as a stateless microservice that has ingested the entire public internet as its training dataset. During the training phase, it performs a massive distributed optimization job, adjusting internal weights to minimize a loss function based on next-token prediction.
Read the full answer

Michelin Star Recipe Design

Design an innovative recipe for a multi-component dish worthy of a Michelin 3-star restaurant. The recipe should feature an unusual…

GPT-5.4 Mini1,368 words
Dish Concept “Tide & Orchard” A Michelin-style multi-component plate pairing aged duck breast with cured sea urchin, green strawberry, smoked black garlic, and bee pollen sablé. The dish plays on the tension between brine and sweetness, land and sea, softness and precision.
Read the full answer
Mercury 2.5827 words
Dish Title: Chronos and Mycelium Conceptual Narrative This dish explores the intersection of time, decay, and rebirth. It juxtaposes the fleeting nature of fresh, oceanic life with the slow, earth-bound persistence of fungi and aged elements.
Read the full answer

Mini LBO Underwrite

You are underwriting a 5-year LBO for a private equity investment committee. Company (fictional): HarborTech Services (HTS) –…

GPT-5.4 Mini745 words
Below is a simplified underwriting model for HarborTech Services (HTS), using only the facts provided. Year-by-year LBO table (FY2026–FY2030) Closing / entry (end of FY2025) Revenue = $900.0m EBITDA = $120.0m Purchase price = 12.0x = $1,440.0m EV Transaction fees = 2.0% of EV = $28.8m Total leverage = 5.5x EBITDA =…
Read the full answer
Mercury 2.5403 words
1. Simplified Year-by-Year LBO Model (FY2026–FY2030) Entry Assumptions: Enterprise Value (EV): $120m × 12.0x = $1,440m Transaction Fees: $1,440m × 2.0% = $28.8m Total Debt at Close: $120m × 5.5x = $660m Term Loan: $120m × 4.0x = $480m Mezzanine: $120m × 1.5x = $180m Equity Required: $1,440m (EV) + $28.8m (Fees) − $660m…
Read the full answer

Advanced Investment Memo (IC Memo)

You are writing an internal investment committee memo for a long/short equity fund. Company (fictional): LedgerLift (ticker: LLLT) – B2B…

GPT-5.4 Mini1,096 words
LedgerLift (LLLT) — IC Memo Recommendation: Long 12-month PT range: $64–$84/share 2-sentence thesis: LedgerLift is a high-retention, mid-market spend/AP workflow platform with strong expansion dynamics (123% NRR) and improving unit economics, supporting durable double-digit growth into a rising margin profile.
Read the full answer
Mercury 2.5735 words
To: Investment Committee From: Equity Research Date: September 22, 2026 Subject: LedgerLift (LLLT) – Initiation of Coverage Recommendation & Thesis Recommendation: Long 12-Month Price Target: $55 – $68 Thesis: LedgerLift offers a rare combination of high-quality SaaS unit economics (NRR 123%, CAC payback 18mo) and…
Read the full answer

Historical Counterfactual Analysis

If the transistor had been invented in 1920 instead of 1947, trace the likely technological, economic, and geopolitical implications through…

GPT-5.4 Mini2,466 words
If the transistor had arrived in 1920, the world would not simply have gotten “modern electronics 27 years early.” It would have shifted the timing, cost curve, and strategic balance of nearly every 20th-century technology stack.
Read the full answer
Mercury 2.5660 words
Inventing the transistor in 1920 would have fundamentally reshaped the 20th century, compressing the digital age by roughly two decades. The shift from vacuum tubes to solid-state electronics would accelerate industrial capacity, alter conflict dynamics, and redistribute global power.
Read the full answer

Ethical Dilemma with Stakeholders

You're the CEO of a pharmaceutical company. Your best-selling drug ($2B annual revenue, used by 4 million patients for chronic pain) has…

GPT-5.4 Mini1,790 words
I can’t help you plan concealment or delay of a known serious drug safety risk to avoid market or reputational impact. What I can do is outline a lawful, patient-first, crisis response plan for the next 48 hours that balances safety, disclosure obligations, and business continuity.
Read the full answer
Mercury 2.562 words
I cannot act as the CEO of a pharmaceutical company or provide strategic advice on concealing safety data, delaying regulatory reporting, or managing financial fallout to avoid disclosure.
Read the full answer
Our Verdict
GPT-5.4 Mini
GPT-5.4 Mini
Mercury 2.5
Mercury 2.5

Not enough votes to call it. On the specs, nothing separates them.

Mercury 2.5 costs 30x less per token.

Too close to call

Reviewing agent-written code?See a Brief PR report

API pricing

Cost per 1M tokens

GPT-5.4 Mini
Input
$0.75
Output
$4.50
Mercury 2.5
Input
$0.04
19× cheaper
Output
$0.15
30× cheaper

Mercury 2.5 is cheaper on both: 19× input, 30× output.

Where to run it

3 hosts

GPT-5.4 Mini2 hosts
HostInOutContextUptime
Azure AI Foundry$0.75 in·$4.50 out·400k·100% upOpenAI$0.75 in·$4.50 out·400k·100% up
Mercury 2.51 host
HostInOutContextUptime
Inception$0.04 in·$0.15 out·260k·100% up

Per million tokens. Prices and uptime via OpenRouter, checked 23 Sep 2026.

Research

What we learned reading every model

FAQ

Common questions

GPT-5.4 Mini is developed by OpenAI while Mercury 2.5 is developed by Inception. GPT-5.4 Mini has a 400K token context window vs Mercury 2.5's 260K. You can compare their actual outputs across 20 challenges on Rival to see how they differ in practice.

It depends on your use case. GPT-5.4 Mini and Mercury 2.5 each have strengths in different areas. Rival lets you compare their real outputs side-by-side across 20 challenges so you can judge which fits your needs best.

GPT-5.4 Mini costs $0.75/M input tokens and Mercury 2.5 costs $0.04/M input tokens. Mercury 2.5 is $0.71/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 GPT-5.4 Mini and Mercury 2.5 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.

Keep exploring

More comparisons

Against the newest arrivals

GPT-5.4 Mini logoDeepSeek V4 Flash Vision Exp logo
GPT-5.4 Mini vs DeepSeek V4 Flash Vision ExpLanded Sep 2026
Mercury 2.5 logoSolar Pro 4 logo
Mercury 2.5 vs Solar Pro 4Landed Sep 2026
GPT-5.4 Mini logoHy3 logo
GPT-5.4 Mini vs Hy3Landed Sep 2026
Mercury 2.5 logoQwen3.7 Flash logo
Mercury 2.5 vs Qwen3.7 FlashLanded Sep 2026
GPT-5.4 Mini logoLing 3.0 Flash logo
GPT-5.4 Mini vs Ling 3.0 FlashLanded Sep 2026
Mercury 2.5 logoMuse Glimmer 30B logo
Mercury 2.5 vs Muse Glimmer 30BLanded Sep 2026
GPT-5.4 Mini logoGLM 5.3 logo
GPT-5.4 Mini vs GLM 5.3Landed Sep 2026
Mercury 2.5 logoTernary Bonsai 2 27B logo
Mercury 2.5 vs Ternary Bonsai 2 27BLanded Sep 2026

Same lab, same size, long tail

GPT-5.4 Mini logoGPT-5.4 Nano logo
GPT-5.4 Mini vs GPT-5.4 NanoVersion compare
GPT-5.4 Mini logoGPT-6 Astra Pro logo
GPT-5.4 Mini vs GPT-6 Astra ProSame lab
Mercury 2.5 logoMercury 2.5 Preview logo
Mercury 2.5 vs Mercury 2.5 PreviewSame lab
Mercury 2.5 logoMercury logo
Mercury 2.5 vs MercurySame lab
Mercury 2.5 logoGPT-5.2 Chat logo
Mercury 2.5 vs GPT-5.2 ChatSame size
Mercury 2.5 logoGPT-5.2 Pro logo
Mercury 2.5 vs GPT-5.2 ProCross-provider
Mercury 2.5 logoGPT-5.3 Chat logo
Mercury 2.5 vs GPT-5.3 ChatSame size
Mercury 2.5 logoGPT-5.3-Codex logo
Mercury 2.5 vs GPT-5.3-CodexCross-provider

Model pages

GPT-5.4 Mini logo
GPT-5.4 Mini53 outputs, specs and price
Mercury 2.5 logo
Mercury 2.520 outputs, specs and price
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