---
title: "Cheaper alternatives to Gemma 2 27B · Undominated.ai"
canonical: https://undominated.ai/alternatives/google__gemma-2-27b-it/
description: "73 models score at least as high as Gemma 2 27B and cost no more on a balanced workload. 73 give up nothing measurable; the other 0 name what they drop. Measured prices and independent scores."
---

# Cheaper alternatives to Gemma 2 27B · Undominated.ai

> 73 models score at least as high as Gemma 2 27B and cost no more on a balanced workload. 73 give up nothing measurable; the other 0 name what they drop. Measured prices and independent scores.

# Cheaper alternatives to Gemma 2 27B

Google · LMArena 1231.5 · $0.650/M on a balanced workload · prices as of 2026-08-28

## 73 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, **73** score at least as high as Gemma 2 27B and cost no more under at least one workload. **73** match or beat it on every capability we hold — context, maximum output, input modes, tool use and extended reasoning.

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 Gemma 2 27B, it is not generated.

## Where it holds

| Workload | Gemma 2 27B $/M | Better and cheaper | With a trade |
| --- | --- | --- | --- |
| Balanced | $0.650 | 50 | 0 |
| Summarise | $0.650 | 72 | 0 |
| Chat | $0.650 | 47 | 0 |
| Code gen | $0.650 | 32 | 0 |
| Agentic | $0.650 | 67 | 0 |

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 | 1231.5 |
| --- | --- |
| Context window | 8K |
| Max output | 2K |
| Input modes | text |
| Tool use | no |
| Extended reasoning | no |

A replacement has to clear every line above, not just the score. [Full record for Gemma 2 27B](/models/google__gemma-2-27b-it/).

## Better and cheaper, nothing given up 73

Each of these matches or beats Gemma 2 27B 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 | MiMo-V2.5-Pro Xiaomi | 1465 +233.5 | $0.544 | Balanced −16% Summarise −49% Chat −26% Agentic −63% |
| 2 | Qwen3.7 Plus Qwen | 1456.2 +224.7 | $0.560 | Balanced −14% Summarise −55% Chat −4% Agentic −52% |
| 3 | Gemma 4 31B Google | 1441.7 +210.2 | $0.152 | Balanced −77% Summarise −86% Chat −73% Code gen −64% Agentic −84% |
| 4 | Hy3 Tencent | 1441.2 +209.7 | $0.144 | Balanced −78% Summarise −88% Chat −75% Code gen −65% Agentic −87% |
| 5 | Gemma 4 26B A4B Google | 1434.6 +203.1 | $0.138 | Balanced −79% Summarise −87% Chat −73% Code gen −64% Agentic −83% |
| 6 | DeepSeek V4 Flash 0423 DeepSeek | 1431.6 +200.1 | $0.101 | Balanced −84% Summarise −90% Chat −86% Code gen −81% Agentic −92% |
| 7 | MiniMax M3 MiniMax | 1434.8 +203.3 | $0.525 | Balanced −19% Summarise −57% Chat −10% Agentic −55% |
| 8 | MiMo-V2.5 Xiaomi | 1427.3 +195.8 | $0.175 | Balanced −73% Summarise −83% Chat −76% Code gen −67% Agentic −88% |
| 9 | GPT-5.6 Luna OpenAI | 1428.5 +197 | $0.450 | Balanced −31% Summarise −69% Chat −16% Agentic −63% |
| 10 | DeepSeek V3.2 DeepSeek | 1424.6 +193.1 | $0.290 | Balanced −55% Summarise −65% Chat −59% Code gen −51% Agentic −69% |
| 11 | DeepSeek V3.2 Exp DeepSeek | 1424.4 +192.9 | $0.305 | Balanced −53% Summarise −57% Chat −50% Code gen −46% Agentic −55% |
| 12 | Qwen3 235B A22B Instruct 2507 Qwen | 1419.3 +187.8 | $0.205 | Balanced −68% Summarise −83% Chat −58% Code gen −44% Agentic −76% |
| 13 | Qwen3 Next 80B A3B Instruct Qwen | 1418.6 +187.1 | $0.350 | Balanced −46% Summarise −78% Chat −24% Agentic −64% |
| 14 | DeepSeek V3.1 Terminus DeepSeek | 1419.6 +188.1 | $0.453 | Balanced −30% Summarise −59% Chat −20% Agentic −54% |
| 15 | Qwen3 VL 235B A22B Instruct Qwen | 1420.9 +189.4 | $0.632 | Balanced −3% Summarise −60% Agentic −39% |
| 16 | Gemini 3.1 Flash Lite Preview Google | 1414.8 +183.3 | $0.563 | Balanced −13% Summarise −62% Agentic −53% |
| 17 | Inkling Small Thinking Machines | 1411.7 +180.2 | $0.637 | Balanced −2% Summarise −40% Chat −1% Agentic −45% |
| 18 | Step 3.5 Flash StepFun | 1403.8 +172.3 | $0.150 | Balanced −77% Summarise −83% Chat −72% Code gen −66% Agentic −80% |
| 19 | Qwen3.5-27B Qwen | 1407.9 +176.4 | $0.536 | Balanced −17% Summarise −60% Agentic −39% |
| 20 | MiniMax M2.7 MiniMax | 1405.3 +173.8 | $0.525 | Balanced −19% Summarise −57% Chat −10% Agentic −55% |
| 21 | Qwen3.5-Flash Qwen | 1397.6 +166.1 | $0.114 | Balanced −82% Summarise −89% Chat −78% Code gen −72% Agentic −85% |
| 22 | Qwen3.5-35B-A3B Qwen | 1395.6 +164.1 | $0.500 | Balanced −23% Summarise −54% Chat same price Agentic −38% |
| 23 | Qwen3 30B A3B Instruct 2507 Qwen | 1384.3 +152.8 | $0.084 | Balanced −87% Summarise −91% Chat −84% Code gen −79% Agentic −89% |
| 24 | GLM 4.5 Air Z.ai | 1382.8 +151.3 | $0.310 | Balanced −52% Summarise −79% Chat −41% Code gen −15% Agentic −73% |
| 25 | Solar Pro 4 Upstage | 1376.2 +144.7 | $0.052 | Balanced −92% Summarise −96% Chat −91% Code gen −87% Agentic −96% |
| 26 | DeepSeek V3 0324 DeepSeek | 1375 +143.5 | $0.438 | Balanced −33% Summarise −56% Chat −15% Agentic −44% |
| 27 | GLM 4.6V Z.ai | 1374.7 +143.2 | $0.450 | Balanced −31% Summarise −60% Chat −28% Code gen −1% Agentic −62% |
| 28 | GPT-5.4 Nano OpenAI | 1372.8 +141.3 | $0.463 | Balanced −29% Summarise −69% Chat −13% Agentic −61% |
| 29 | gpt-oss-120b OpenAI | 1365.6 +134.1 | $0.070 | Balanced −89% Summarise −93% Chat −86% Code gen −82% Agentic −91% |
| 30 | Qwen3 Next 80B A3B Thinking Qwen | 1367.5 +136 | $0.412 | Balanced −37% Summarise −69% Chat −12% Agentic −53% |
| 31 | Gemma 3 27B Google | 1358.3 +126.8 | $0.172 | Balanced −73% Summarise −87% Chat −67% Code gen −54% Agentic −83% |
| 32 | Mercury 2 Inception | 1357.8 +126.3 | $0.375 | Balanced −42% Summarise −68% Chat −41% Code gen −18% Agentic −71% |
| 33 | MiniMax M2.5 MiniMax | 1359 +127.5 | $0.472 | Balanced −27% Summarise −63% Chat −20% Agentic −62% |
| 34 | GLM 4.7 Flash Z.ai | 1352.9 +121.4 | $0.145 | Balanced −78% Summarise −90% Chat −72% Code gen −60% Agentic −87% |
| 35 | Qwen3 32B Qwen | 1340.1 +108.6 | $0.130 | Balanced −80% Summarise −86% Chat −75% Code gen −69% Agentic −83% |
| 36 | Trinity Large Thinking Arcee | 1341.9 +110.4 | $0.378 | Balanced −42% Summarise −68% Chat −35% Code gen −10% Agentic −66% |
| 37 | MiniMax M2 MiniMax | 1342.1 +110.6 | $0.446 | Balanced −31% Summarise −55% Chat −14% Agentic −43% |
| 38 | Gemma 3 12B Google | 1334.2 +102.7 | $0.075 | Balanced −88% Summarise −92% Chat −86% Code gen −83% Agentic −90% |
| 39 | DeepSeek V3 DeepSeek | 1332.6 +101.1 | $0.450 | Balanced −31% Summarise −54% Chat −13% Agentic −43% |
| 40 | GPT-5 Nano OpenAI | 1320.3 +88.8 | $0.138 | Balanced −79% Summarise −92% Chat −73% Code gen −61% Agentic −88% |
| 41 | Qwen3 30B A3B Qwen | 1316.9 +85.4 | $0.215 | Balanced −67% Summarise −79% Chat −58% Code gen −46% Agentic −73% |
| 42 | Olmo 3 32B Think Allen AI | 1298.5 +67 | $0.237 | Balanced −63% Summarise −74% Chat −55% Code gen −45% Agentic −69% |
| 43 | Granite 4.1 8B IBM | 1291.6 +60.1 | $0.063 | Balanced −90% Summarise −92% Chat −89% Code gen −88% Agentic −91% |
| 44 | Gemma 3 4B Google | 1290.8 +59.3 | $0.063 | Balanced −90% Summarise −92% Chat −89% Code gen −88% Agentic −91% |
| 45 | gpt-oss-20b OpenAI | 1287.8 +56.3 | $0.055 | Balanced −92% Summarise −95% Chat −89% Code gen −86% Agentic −93% |
| 46 | GPT-4o-mini (2024-07-18) OpenAI | 1286.6 +55.1 | $0.263 | Balanced −60% Summarise −77% Chat −53% Code gen −36% Agentic −73% |
| 47 | GPT-4.1 Nano OpenAI | 1284.8 +53.3 | $0.175 | Balanced −73% Summarise −86% Chat −70% Code gen −58% Agentic −85% |
| 48 | Qwen2.5 72B Instruct Qwen | 1269.1 +37.6 | $0.370 | Balanced −43% Summarise −44% Chat −42% Code gen −41% Agentic −44% |
| 49 | Llama 3.1 70B Instruct Meta | 1261 +29.5 | $0.400 | Balanced −38% Summarise −38% Chat −38% Code gen −38% Agentic −38% |
| 50 | Mistral Small 3 Mistral | 1233.6 tie | $0.058 | Balanced −91% Summarise −92% Chat −90% Code gen −90% Agentic −92% |
| 51 | Gemini 3.7 Flash Google | 1490.2 +258.7 | $0.354 summarise | Summarise −46% Agentic −39% |
| 52 | Kimi K2.5 Moonshot AI | 1445.2 +213.7 | $0.432 summarise | Summarise −34% Agentic −24% |
| 53 | GLM 5 Z.ai | 1445.2 +213.7 | $0.529 summarise | Summarise −19% Agentic −21% |
| 54 | Qwen3.8 27B Qwen | 1440.8 +209.3 | $0.434 summarise | Summarise −33% Agentic −17% |
| 55 | GLM 4.6 Z.ai | 1439.8 +208.3 | $0.461 summarise | Summarise −29% Agentic −25% |
| 56 | Gemini 3.5 Flash Lite Google | 1436.5 +205 | $0.333 summarise | Summarise −49% Agentic −28% |
| 57 | DeepSeek V4 Pro 0423 DeepSeek | 1439.2 +207.7 | $0.526 agentic | Agentic −19% |
| 58 | Qwen3.5 397B A17B Qwen | 1438.3 +206.8 | $0.487 summarise | Summarise −25% |
| 59 | Qwen3.6 Plus Qwen | 1436.8 +205.3 | $0.406 summarise | Summarise −37% Agentic −13% |
| 60 | GLM 4.7 Z.ai | 1435.3 +203.8 | $0.376 summarise | Summarise −42% Agentic −37% |
| 61 | GLM 4.5 Z.ai | 1429.4 +197.9 | $0.540 summarise | Summarise −17% Agentic −16% |
| 62 | R1 0528 DeepSeek | 1427.9 +196.4 | $0.540 summarise | Summarise −17% |
| 63 | Qwen3.5-122B-A10B Qwen | 1417.9 +186.4 | $0.351 summarise | Summarise −46% Agentic −18% |
| 64 | Gemini 2.5 Flash Google | 1417.3 +185.8 | $0.333 summarise | Summarise −49% Agentic −28% |
| 65 | DeepSeek V3.1 DeepSeek | 1419.1 +187.6 | $0.605 summarise | Summarise −7% |
| 66 | Qwen3 235B A22B Thinking 2507 Qwen | 1413.8 +182.3 | $0.334 summarise | Summarise −49% Agentic −17% |
| 67 | Qwen3 VL 235B A22B Thinking Qwen | 1400.6 +169.1 | $0.580 summarise | Summarise −11% |
| 68 | GPT-5 Mini OpenAI | 1373.4 +141.9 | $0.273 summarise | Summarise −58% Agentic −42% |
| 69 | Qwen3 235B A22B Qwen | 1366 +134.5 | $0.523 summarise | Summarise −20% |
| 70 | Nova 2 Lite Amazon | 1362.2 +130.7 | $0.410 summarise | Summarise −37% Agentic −3% |
| 71 | GPT-4.1 Mini OpenAI | 1340.5 +109 | $0.374 summarise | Summarise −42% Agentic −38% |
| 72 | MiniMax M1 MiniMax | 1341.9 +110.4 | $0.632 summarise | Summarise −3% |
| 73 | GLM 4.5V Z.ai | 1333.6 +102.1 | $0.520 summarise | Summarise −20% Agentic −25% |

## 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/).