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. Gemma 4 26B A4BvsMercury 2.5
Updated Sep 2026

Gemma 4 26B A4BvsMercury 2.5

Mercury 2.5 is cheaper than Gemma 4 26B A4B at $0.04/M vs $0.13/M input tokens.

Gemma 4 26B A4B and Mercury 2.5 compared across 20 shared prompts
SpecGemma 4 26B A4BMercury 2.5
Input price$0.13/M tokens$0.04/M tokens
Output price$0.4/M tokens$0.15/M tokens
Context window262K tokens260K tokens
WeightsOpen—
Free API (OpenRouter)Yes (1 provider)No
ReleasedApr 2026Sep 2026
Side by side

Same prompt, both answers

Opening lines from 6 of the 20 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…

Gemma 4 26B A4B735 words
To trace this counterfactual, we must assume that the fundamental physics of solid-state semiconductors (the ability to control electron flow through crystals) was mastered during the vacuum tube era. This shifts the "Information Age" forward by nearly 30 years. Here is the timeline of implications from 1920 to 1980.
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

Startup Pitch Teardown

Here's a pitch deck summary for a startup called "MindMeld AI": **Slide 1 - Vision:** "We're building the future of human-AI collaboration.…

Gemma 4 26B A4B558 words
This is a high-ambition pitch that falls into a common trap for deep-tech startups: the "Science Fiction Gap." While the vision is compelling, several claims lack the scientific and operational credibility required to justify a $15M Series A.
Read the full answer
Mercury 2.5544 words
Based on the pitch deck summary, here are the three weakest claims, along with an analysis of their vulnerabilities and concrete suggestions for improvement. The "Pre-Thought" Prediction Claim Claim: "MindMeld AI reads your brainwaves to predict what you want to type before you think it."
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) –…

Gemma 4 26B A4B435 words
Financial Model (FY2026–FY2030) Entry Calculations (End of FY2025): Purchase Price (EV): $120\text{m} \times 12.0\text{x} = \$1,440\text{m}$ Transaction Fees: $1,440\text{m} \times 2\% = \$28.8\text{m}$ Total Uses: $\$1,440\text{m} + \$28.8\text{m} = \$1,468.8\text{m}$ Debt (Total 5.5x): $\$120\text{m} \times 5.5 =…
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…

Gemma 4 26B A4B363 words
Here is a recipe for a comforting, savory dish that relies on pantry staples to create a deep, rich flavor. The Golden Pantry Pasta A silky, garlic-infused pasta dish with a crispy, salty crunch.
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

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…

Gemma 4 26B A4B699 words
Concept Narrative: "The Forest Floor & The Ocean’s Breath" The Concept: This dish explores the tension between the terrestrial and the maritime. We are pairing the deep, earthy, almost primal musk of Black Truffle with the delicate, saline, and floral notes of Sea Urchin (Uni).
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

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…

Gemma 4 26B A4B624 words
To the Experienced Software Engineer Focus: Complexity, Emergent Properties, and the "Training Pipeline" Think of an LLM not as a magical oracle, but as a massive, non-deterministic state machine trained via a high-throughput distributed training pipeline.
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
Our Verdict
Gemma 4 26B A4B
Gemma 4 26B A4B
Mercury 2.5
Mercury 2.5

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

Too close to call

Reviewing agent-written code?See a Brief PR report

API pricing

Cost per 1M tokens

Gemma 4 26B A4B
Input
$0.13
Output
$0.40
Mercury 2.5
Input
$0.04
3.3× cheaper
Output
$0.15
2.7× cheaper

Mercury 2.5 is cheaper on both: 3.3× input, 2.7× output.

Where to run it

13 hosts, cheapest first

Gemma 4 26B A4B12 hosts
HostInOutContextUptime
DDarkbloom$0.04 in·$0.22 out·131k·99.9% upDDekaLLMbf16$0.06 in·$0.33 out·262k·99.9% upDDeepInfrafp8$0.07 in·$0.34 out·262k·99.4% upNNextBitbf16$0.09 in·$0.30 out·262k·97.3% upCCoreWeavebf16$0.10 in·$0.30 out·262k·98.5% upMMakora$0.10 in·$0.34 out·262k·99% up
6 more hostsFewer hosts
NNovitabf16$0.13 in·$0.40 out·262k·99.3% upPParasailbf16$0.13 in·$0.40 out·262k·98.8% upVVenicebf16$0.13 in·$0.40 out·256k·97.6% upSSiliconFlowfp8$0.14 in·$0.40 out·262k·97.7% upGoogle Vertex AI$0.15 in·$0.60 out·262k·96% upCloudflare Workers AIdegraded$0.10 in·$0.30 out·256k·99% 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

Gemma 4 26B A4B is developed by Google AI while Mercury 2.5 is developed by Inception. Gemma 4 26B A4B has a 262K 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. Gemma 4 26B A4B 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.

Gemma 4 26B A4B costs $0.13/M input tokens and Mercury 2.5 costs $0.04/M input tokens. Mercury 2.5 is $0.09/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 26B A4B 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

Gemma 4 26B A4B logoDeepSeek V4 Flash Vision Exp logo
Gemma 4 26B A4B vs DeepSeek V4 Flash Vision ExpLanded Sep 2026
Mercury 2.5 logoSolar Pro 4 logo
Mercury 2.5 vs Solar Pro 4Landed Sep 2026
Gemma 4 26B A4B logoHy3 logo
Gemma 4 26B A4B vs Hy3Landed Sep 2026
Mercury 2.5 logoQwen3.7 Flash logo
Mercury 2.5 vs Qwen3.7 FlashLanded Sep 2026
Gemma 4 26B A4B logoLing 3.0 Flash logo
Gemma 4 26B A4B vs Ling 3.0 FlashLanded Sep 2026
Mercury 2.5 logoMuse Glimmer 30B logo
Mercury 2.5 vs Muse Glimmer 30BLanded Sep 2026
Gemma 4 26B A4B logoGLM 5.3 logo
Gemma 4 26B A4B 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

Gemma 4 26B A4B logoGemini 3.8 Flash logo
Gemma 4 26B A4B vs Gemini 3.8 FlashSame lab
Gemma 4 26B A4B logoGemini 3.7 Flash logo
Gemma 4 26B A4B vs Gemini 3.7 FlashSame 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-5.1-Codex logo
Mercury 2.5 vs GPT-5.1-CodexNew provider
Mercury 2.5 logoGPT-5.1 Codex Max logo
Mercury 2.5 vs GPT-5.1 Codex MaxNew provider
Mercury 2.5 logoGPT-5.1-Codex-Mini logo
Mercury 2.5 vs GPT-5.1-Codex-MiniNew provider
Mercury 2.5 logoGPT-5.2 logo
Mercury 2.5 vs GPT-5.2New provider

Model pages

Gemma 4 26B A4B logo
Gemma 4 26B A4B58 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