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 ProvsMercury 2.5
Updated Sep 2026

GPT-5.4 ProvsMercury 2.5

Mercury 2.5 is cheaper than GPT-5.4 Pro at $0.04/M vs $30/M input tokens.

GPT-5.4 Pro and Mercury 2.5 compared across 18 shared prompts
SpecGPT-5.4 ProMercury 2.5
Input price$30/M tokens$0.04/M tokens
Output price$180/M tokens$0.15/M tokens
Context window1.1M tokens260K tokens
WeightsClosed—
Free API (OpenRouter)NoNo
ReleasedMar 2026Sep 2026
Side by side

Same prompt, both answers

Opening lines from 6 of the 18 prompts both answered. Formatting stripped, words untouched.

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 Pro2,183 words
Most likely, a transistor invented in 1920 would move the electronics/computing frontier forward by about 10–15 years by 1980, not the full 27 years. Reason: the transistor alone is not enough; you also need high-purity materials, crystal growth, photolithography, test equipment, software, batteries, precision…
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

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 Pro1,064 words
LedgerLift (LLLT) — IC Memo Recommendation: Pass 12-month PT range: $40–$47 2-sentence thesis: LedgerLift looks like a good business but only an average stock here: retention is strong (94% GRR, 123% NRR), the model is mostly subscription, and margins are inflecting, but at $46 the shares already discount a lot of the…
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

Mini LBO Underwrite

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

GPT-5.4 Pro731 words
Below is a simplified LBO build for HarborTech Services (HTS). All figures in $m unless noted. Rounding: tables are rounded; IRR/MOIC use unrounded math. Entry / Sources & Uses FY2025 EBITDA: 120.0 Purchase price: 12.0x EBITDA → Enterprise value = 12.0 × 120.0 = 1,440.0 Transaction fees: 2.0% of EV → 28.8 Total uses =…
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

Simple 7-Ingredient Recipe

Create a simple recipe for a delicious dish (sweet or savory) using only 7 common pantry ingredients. Include a creative name for the dish…

GPT-5.4 Pro151 words
Dragon Pantry Peanut Noodles A fast, sweet-savory noodle dish made from 7 pantry staples. Ingredients 8 oz spaghetti 1/3 cup peanut butter 3 tbsp soy sauce 1 tbsp honey 1 tbsp vegetable oil 1 tsp garlic powder 1/4 tsp red pepper flakes Water for boiling/thinning the sauce doesn’t count.
Read the full answer
Mercury 2.5190 words
Pantry Pasta Primavera Prep Time: 5 minutes Cook Time: 15 minutes Servings: 2 Ingredients (7 Common Pantry Items) Pasta (spaghetti, penne, or fusilli) – 8 oz Canned crushed tomatoes – 1 can (14 oz) Olive oil – 3 tablespoons Garlic powder – 1 teaspoon Dried oregano – 1 teaspoon Parmesan cheese (grated) – ½ cup Salt and…
Read the full answer

Adversarial Contract Review

Here is a contract for a software consulting engagement: --- SOFTWARE CONSULTING AGREEMENT This Agreement is entered into between TechCorp…

GPT-5.4 Pro1,392 words
This agreement is heavily client-favored. Every numbered clause has language that could be used against the contractor. Short version: I would not sign it without major redlines to Clauses 1–9, especially 2, 3, 4, 5, 6, and 7. Below is a clause-by-clause review.
Read the full answer
Mercury 2.5501 words
Here is an analysis of the contract clauses that pose significant risk to the contractor, along with suggested modifications and legal reasoning. Scope and Compensation (Clause 1) Risk: The client can expand the project indefinitely without paying extra.
Read the full answer

Debug This Architecture

Here's the architecture for a real-time collaborative document editor (like Google Docs): **System Overview:** - Frontend: React SPA with…

GPT-5.4 Pro2,331 words
This architecture will work for a demo, but it has major correctness, latency, and security problems for a Google-Docs-like editor. Highest-risk flaws Client-clock last-write-wins will lose edits and can be gamed by bad clocks.
Read the full answer
Mercury 2.5672 words
Critical Analysis of Collaborative Editor Architecture Below is a detailed breakdown of failure modes, race conditions, and scaling bottlenecks, along with proposed solutions and their trade-offs.
Read the full answer
Our Verdict
GPT-5.4 Pro
GPT-5.4 Pro
Mercury 2.5
Mercury 2.5Runner-up

Not enough votes to call it. On the specs, GPT-5.4 Pro has the edge: bigger model tier, bigger context window, major provider backing.

Mercury 2.5 costs 1200x 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 Pro
Input
$30.00
Output
$180.00
Mercury 2.5
Input
$0.04
750× cheaper
Output
$0.15
1200× cheaper

Mercury 2.5 is cheaper on both: 750× input, 1200× output.

Where to run it

3 hosts

GPT-5.4 Pro2 hosts
HostInOutContextUptime
Azure AI Foundry$30.00 in·$180.00 out·1.1M—OpenAI$30.00 in·$180.00 out·1.1M·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 Pro is developed by OpenAI while Mercury 2.5 is developed by Inception. GPT-5.4 Pro has a 1.1M token context window vs Mercury 2.5's 260K. You can compare their actual outputs across 18 challenges on Rival to see how they differ in practice.

It depends on your use case. GPT-5.4 Pro and Mercury 2.5 each have strengths in different areas. Rival lets you compare their real outputs side-by-side across 18 challenges so you can judge which fits your needs best.

GPT-5.4 Pro costs $30/M input tokens and Mercury 2.5 costs $0.04/M input tokens. Mercury 2.5 is $29.96/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 Pro 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 Pro logoDeepSeek V4 Flash Vision Exp logo
GPT-5.4 Pro 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 Pro logoHy3 logo
GPT-5.4 Pro vs Hy3Landed Sep 2026
Mercury 2.5 logoQwen3.7 Flash logo
Mercury 2.5 vs Qwen3.7 FlashLanded Sep 2026
GPT-5.4 Pro logoLing 3.0 Flash logo
GPT-5.4 Pro 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 Pro logoGLM 5.3 logo
GPT-5.4 Pro 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 Pro logoGPT-5.4 logo
GPT-5.4 Pro vs GPT-5.4Version compare
GPT-5.4 Pro logoGPT-6 Astra logo
GPT-5.4 Pro vs GPT-6 AstraSame 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 logoMuse Spark 1.3 Contributor logo
Mercury 2.5 vs Muse Spark 1.3 ContributorSame size
Mercury 2.5 logoNemotron 3 Nano Omni (free) logo
Mercury 2.5 vs Nemotron 3 Nano Omni (free)Same size
Mercury 2.5 logoNVIDIA Nemotron 3 Super (free) logo
Mercury 2.5 vs NVIDIA Nemotron 3 Super (free)Same size
Mercury 2.5 logoNemotron 3 Ultra logo
Mercury 2.5 vs Nemotron 3 UltraNew provider

Model pages

GPT-5.4 Pro logo
GPT-5.4 Pro18 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