---
title: "Cheaper alternatives to Mercury 2 · Undominated.ai"
canonical: https://undominated.ai/alternatives/inception__mercury-2/
description: "19 models score at least as high as Mercury 2 and cost no more on a balanced workload. 12 give up nothing measurable; the other 7 name what they drop. Measured prices and independent scores."
---

# Cheaper alternatives to Mercury 2 · Undominated.ai

> 19 models score at least as high as Mercury 2 and cost no more on a balanced workload. 12 give up nothing measurable; the other 7 name what they drop. Measured prices and independent scores.

# Cheaper alternatives to Mercury 2

Inception · LMArena 1357.8 · $0.375/M on a balanced workload · prices as of 2026-08-28

## 12 models are both better and cheaper, giving up nothing.

Of the 133 models carrying both an independent score and a published price at the same delivery mode, **19** score at least as high as Mercury 2 and cost no more under at least one workload. **12** match or beat it on every capability we hold — context, maximum output, input modes, tool use and extended reasoning. **7** do not, and every row names what it drops.

Dominance is not a property of a model. It is a property of a model, a capability lens and a workload, and all three are named on every row. Scores are LMArena Elo, used under CC BY 4.0 from the official dataset; prices are blended per million tokens. A higher score is not a drop-in replacement.

This page is built from that comparison and nothing else. When no model is both better and cheaper than Mercury 2, it is not generated.

## Where it holds

| Workload | Mercury 2 $/M | Better and cheaper | With a trade |
| --- | --- | --- | --- |
| Balanced | $0.375 | 10 | 6 |
| Summarise | $0.211 | 10 | 7 |
| Chat | $0.383 | 9 | 5 |
| Code gen | $0.532 | 9 | 5 |
| Agentic | $0.191 | 8 | 5 |

Workload mixes are defined on the [methodology page](/methodology/). Cached input is priced at the cached rate, and reasoning tokens at the worse of the reasoning and output rates.

## What you would be replacing

| Intelligence | 1357.8 |
| --- | --- |
| Context window | 128K |
| Max output | 50K |
| Input modes | text |
| Tool use | yes |
| Extended reasoning | yes |

A replacement has to clear every line above, not just the score. [Full record for Mercury 2](/models/inception__mercury-2/).

## Better and cheaper, nothing given up 12

Each of these matches or beats Mercury 2 on context, maximum output, input modes, tool use and reasoning, scores at least as high, and costs no more.

| # | Model | Intelligence | $/M | Holds under |
| --- | --- | --- | --- | --- |
| 1 | Hy3 Tencent | 1441.2 +83.4 | $0.144 | Balanced −61% Summarise −63% Chat −57% Code gen −58% Agentic −57% |
| 2 | DeepSeek V4 Flash 0423 DeepSeek | 1431.6 +73.8 | $0.101 | Balanced −73% Summarise −68% Chat −75% Code gen −77% Agentic −71% |
| 3 | MiMo-V2.5 Xiaomi | 1427.3 +69.5 | $0.175 | Balanced −53% Summarise −49% Chat −60% Code gen −60% Agentic −58% |
| 4 | DeepSeek V3.2 DeepSeek | 1424.6 +66.8 | $0.290 | Balanced −23% Chat −30% Code gen −40% |
| 5 | DeepSeek V3.2 Exp DeepSeek | 1424.4 +66.6 | $0.305 | Balanced −19% Chat −15% Code gen −33% |
| 6 | Step 3.5 Flash StepFun | 1403.8 +46 | $0.150 | Balanced −60% Summarise −48% Chat −53% Code gen −59% Agentic −32% |
| 7 | Qwen3.5-Flash Qwen | 1397.6 +39.8 | $0.114 | Balanced −70% Summarise −65% Chat −63% Code gen −66% Agentic −51% |
| 8 | Solar Pro 4 Upstage | 1376.2 +18.4 | $0.052 | Balanced −86% Summarise −87% Chat −85% Code gen −85% Agentic −85% |
| 9 | GLM 4.5 Air Z.ai | 1382.8 +25 | $0.310 | Balanced −17% Summarise −35% Agentic −8% |
| 10 | gpt-oss-120b OpenAI | 1365.6 tie | $0.070 | Balanced −81% Summarise −79% Chat −76% Code gen −78% Agentic −70% |
| 11 | GPT-5.6 Luna OpenAI | 1428.5 +70.7 | $0.199 summarise | Summarise −6% |
| 12 | GPT-5.4 Nano OpenAI | 1372.8 +15 | $0.201 summarise | Summarise −5% |

## Cheaper and higher-scoring, but you give something up 7

These score at least as high and cost no more on the two plotted axes, and lose something that is not on them. Read the last column before switching.

| Model | Intelligence | $/M | Holds under | What you give up |
| --- | --- | --- | --- | --- |
| Gemma 4 31B Google | 1441.7 +83.9 | $0.152 | Balanced −59% Summarise −57% Chat −53% Code gen −55% Agentic −46% | 50K → 16K max output |
| Gemma 4 26B A4B Google | 1434.6 +76.8 | $0.138 | Balanced −63% Summarise −60% Chat −53% Code gen −56% Agentic −42% | 50K → 16K max output |
| Qwen3 235B A22B Instruct 2507 Qwen | 1419.3 +61.5 | $0.205 | Balanced −45% Summarise −46% Chat −28% Code gen −31% Agentic −17% | 50K → 16K max output no extended reasoning |
| Qwen3 Next 80B A3B Instruct Qwen | 1418.6 +60.8 | $0.350 | Balanced −7% Summarise −33% | no extended reasoning |
| Qwen3 30B A3B Instruct 2507 Qwen | 1384.3 +26.5 | $0.084 | Balanced −77% Summarise −74% Chat −72% Code gen −75% Agentic −63% | 50K → 32K max output no extended reasoning |
| Gemma 3 27B Google | 1358.3 tie | $0.172 | Balanced −54% Summarise −59% Chat −44% Code gen −44% Agentic −42% | no extended reasoning |
| Qwen3 Next 80B A3B Thinking Qwen | 1367.5 tie | $0.203 summarise | Summarise −4% | 50K → 33K max output |

## What this compares, and what it leaves out

 - Quality is LMArena Elo, used under CC BY 4.0 from the official dataset. The 95% confidence interval on a difference between two scores is about ±10.68 points, so a gap smaller than that is marked *tie* rather than an improvement — see [significance bands](/significance/).
 - 188 further models at this delivery mode carry a price but no independent score. They are absent from the comparison in both directions — unrated is not a zero, and an unmeasured model is neither an alternative nor a worse buy.
 - Retired models are never offered as an alternative, and a model only competes against its own delivery mode: batch trades latency for price, so it is not a like-for-like swap.
 - Nothing here measures latency, throughput, rate limits or how a model behaves on your prompts. Two models with the same index score are not interchangeable.
 - The models that nothing beats on both axes are on the [value frontier](/frontier/), and every other model something cheaper beats is [listed here](/alternatives/).