---
title: "Cheaper alternatives to R1 · Undominated.ai"
canonical: https://undominated.ai/alternatives/deepseek__deepseek-r1/
description: "52 models score at least as high as R1 and cost no more on a balanced workload. 47 give up nothing measurable; the other 5 name what they drop. Measured prices and independent scores."
---

# Cheaper alternatives to R1 · Undominated.ai

> 52 models score at least as high as R1 and cost no more on a balanced workload. 47 give up nothing measurable; the other 5 name what they drop. Measured prices and independent scores.

# Cheaper alternatives to R1

DeepSeek · LMArena 1372.7 · $1.15/M on a balanced workload · prices as of 2026-08-28

## 47 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, **52** score at least as high as R1 and cost no more under at least one workload. **47** match or beat it on every capability we hold — context, maximum output, input modes, tool use and extended reasoning. **5** 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 R1, it is not generated.

## Where it holds

| Workload | R1 $/M | Better and cheaper | With a trade |
| --- | --- | --- | --- |
| Balanced | $1.15 | 40 | 5 |
| Summarise | $0.790 | 44 | 5 |
| Chat | $1.42 | 40 | 5 |
| Code gen | $1.78 | 40 | 5 |
| Agentic | $0.970 | 47 | 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 | 1372.7 |
| --- | --- |
| Context window | 64K |
| Max output | 16K |
| Input modes | text |
| Tool use | yes |
| Extended reasoning | yes |

A replacement has to clear every line above, not just the score. [Full record for R1](/models/deepseek__deepseek-r1/).

## Better and cheaper, nothing given up 47

Each of these matches or beats R1 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 | Gemini 3.7 Flash Google | 1490.2 +117.5 | $0.750 | Balanced −35% Summarise −55% Chat −38% Code gen −30% Agentic −59% |
| 2 | MiMo-V2.5-Pro Xiaomi | 1465 +92.3 | $0.544 | Balanced −53% Summarise −58% Chat −66% Code gen −63% Agentic −75% |
| 3 | Qwen3.7 Plus Qwen | 1456.2 +83.5 | $0.560 | Balanced −51% Summarise −63% Chat −56% Code gen −51% Agentic −68% |
| 4 | Gemma 4 31B Google | 1441.7 +69 | $0.152 | Balanced −87% Summarise −88% Chat −87% Code gen −87% Agentic −89% |
| 5 | Hy3 Tencent | 1441.2 +68.5 | $0.144 | Balanced −87% Summarise −90% Chat −89% Code gen −87% Agentic −91% |
| 6 | Kimi K2.5 Moonshot AI | 1445.2 +72.5 | $0.900 | Balanced −22% Summarise −45% Chat −26% Code gen −16% Agentic −49% |
| 7 | GLM 5 Z.ai | 1445.2 +72.5 | $0.930 | Balanced −19% Summarise −33% Chat −31% Code gen −24% Agentic −47% |
| 8 | Gemma 4 26B A4B Google | 1434.6 +61.9 | $0.138 | Balanced −88% Summarise −89% Chat −87% Code gen −87% Agentic −89% |
| 9 | Qwen3.8 27B Qwen | 1440.8 +68.1 | $0.956 | Balanced −17% Summarise −45% Chat −17% Code gen −6% Agentic −44% |
| 10 | GLM 4.6 Z.ai | 1439.8 +67.1 | $0.875 | Balanced −24% Summarise −42% Chat −31% Code gen −23% Agentic −50% |
| 11 | DeepSeek V4 Flash 0423 DeepSeek | 1431.6 +58.9 | $0.101 | Balanced −91% Summarise −92% Chat −93% Code gen −93% Agentic −94% |
| 12 | Qwen3.5 397B A17B Qwen | 1438.3 +65.6 | $0.877 | Balanced −24% Summarise −38% Chat −18% Code gen −12% Agentic −30% |
| 13 | Qwen3.6 Plus Qwen | 1436.8 +64.1 | $0.731 | Balanced −36% Summarise −49% Chat −31% Code gen −27% Agentic −41% |
| 14 | MiniMax M3 MiniMax | 1434.8 +62.1 | $0.525 | Balanced −54% Summarise −65% Chat −59% Code gen −54% Agentic −70% |
| 15 | DeepSeek V4 Pro 0423 DeepSeek | 1439.2 +66.5 | $1.09 | Balanced −5% Summarise −13% Chat −31% Code gen −25% Agentic −46% |
| 16 | Gemini 3.5 Flash Lite Google | 1436.5 +63.8 | $0.850 | Balanced −26% Summarise −58% Chat −23% Code gen −10% Agentic −52% |
| 17 | GLM 4.7 Z.ai | 1435.3 +62.6 | $0.738 | Balanced −36% Summarise −52% Chat −41% Code gen −33% Agentic −58% |
| 18 | MiMo-V2.5 Xiaomi | 1427.3 +54.6 | $0.175 | Balanced −85% Summarise −86% Chat −89% Code gen −88% Agentic −92% |
| 19 | GPT-5.6 Luna OpenAI | 1428.5 +55.8 | $0.450 | Balanced −61% Summarise −75% Chat −62% Code gen −56% Agentic −75% |
| 20 | DeepSeek V3.2 DeepSeek | 1424.6 +51.9 | $0.290 | Balanced −75% Summarise −71% Chat −81% Code gen −82% Agentic −79% |
| 21 | DeepSeek V3.2 Exp DeepSeek | 1424.4 +51.7 | $0.305 | Balanced −73% Summarise −65% Chat −77% Code gen −80% Agentic −70% |
| 22 | GLM 4.5 Z.ai | 1429.4 +56.7 | $1.00 | Balanced −13% Summarise −32% Chat −23% Code gen −15% Agentic −43% |
| 23 | R1 0528 DeepSeek | 1427.9 +55.2 | $0.912 | Balanced −21% Summarise −32% Chat −21% Code gen −17% Agentic −32% |
| 24 | DeepSeek V3.1 Terminus DeepSeek | 1419.6 +46.9 | $0.453 | Balanced −61% Summarise −66% Chat −63% Code gen −61% Agentic −69% |
| 25 | DeepSeek V3.1 DeepSeek | 1419.1 +46.4 | $0.825 | Balanced −28% Summarise −23% Chat −30% Code gen −32% Agentic −26% |
| 26 | Qwen3.5-122B-A10B Qwen | 1417.9 +45.2 | $0.715 | Balanced −38% Summarise −56% Chat −30% Code gen −24% Agentic −45% |
| 27 | Gemini 2.5 Flash Google | 1417.3 +44.6 | $0.850 | Balanced −26% Summarise −58% Chat −23% Code gen −10% Agentic −52% |
| 28 | Gemini 3.1 Flash Lite Preview Google | 1414.8 +42.1 | $0.563 | Balanced −51% Summarise −69% Chat −52% Code gen −45% Agentic −69% |
| 29 | Qwen3 235B A22B Thinking 2507 Qwen | 1413.8 +41.1 | $0.748 | Balanced −35% Summarise −58% Chat −25% Code gen −17% Agentic −44% |
| 30 | Inkling Small Thinking Machines | 1411.7 +39 | $0.637 | Balanced −45% Summarise −51% Chat −55% Code gen −51% Agentic −63% |
| 31 | Qwen3.5-27B Qwen | 1407.9 +35.2 | $0.536 | Balanced −53% Summarise −67% Chat −48% Code gen −43% Agentic −59% |
| 32 | Step 3.5 Flash StepFun | 1403.8 +31.1 | $0.150 | Balanced −87% Summarise −86% Chat −87% Code gen −88% Agentic −87% |
| 33 | MiniMax M2.7 MiniMax | 1405.3 +32.6 | $0.525 | Balanced −54% Summarise −65% Chat −59% Code gen −54% Agentic −70% |
| 34 | Qwen3.5-Flash Qwen | 1397.6 +24.9 | $0.114 | Balanced −90% Summarise −91% Chat −90% Code gen −90% Agentic −90% |
| 35 | Qwen3.5-35B-A3B Qwen | 1395.6 +22.9 | $0.500 | Balanced −57% Summarise −62% Chat −54% Code gen −52% Agentic −59% |
| 36 | GLM 4.5 Air Z.ai | 1382.8 tie | $0.310 | Balanced −73% Summarise −83% Chat −73% Code gen −69% Agentic −82% |
| 37 | Solar Pro 4 Upstage | 1376.2 tie | $0.052 | Balanced −95% Summarise −96% Chat −96% Code gen −95% Agentic −97% |
| 38 | GLM 4.6V Z.ai | 1374.7 tie | $0.450 | Balanced −61% Summarise −67% Chat −67% Code gen −64% Agentic −75% |
| 39 | GPT-5.4 Nano OpenAI | 1372.8 tie | $0.463 | Balanced −60% Summarise −75% Chat −60% Code gen −54% Agentic −74% |
| 40 | GPT-5 Mini OpenAI | 1373.4 tie | $0.688 | Balanced −40% Summarise −65% Chat −38% Code gen −28% Agentic −61% |
| 41 | Gemini 3.6 Flash Google | 1476.5 +103.8 | $0.708 summarise | Summarise −10% Agentic −18% |
| 42 | Kimi K2.6 Moonshot AI | 1455 +82.3 | $0.938 agentic | Agentic −3% |
| 43 | Inkling Thinking Machines | 1439.2 +66.5 | $0.945 agentic | Agentic −3% |
| 44 | Qwen3 Max Qwen | 1412.7 +40 | $0.758 summarise | Summarise −4% Agentic −10% |
| 45 | GPT-5.4 Mini OpenAI | 1412.1 +39.4 | $0.745 summarise | Summarise −6% Agentic −6% |
| 46 | Qwen3 VL 235B A22B Thinking Qwen | 1400.6 +27.9 | $0.580 summarise | Summarise −27% Agentic −3% |
| 47 | Grok 4.3 xAI | 1397.5 +24.8 | $0.813 agentic | Agentic −16% |

## Cheaper and higher-scoring, but you give something up 5

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 |
| --- | --- | --- | --- | --- |
| Qwen3 235B A22B Instruct 2507 Qwen | 1419.3 +46.6 | $0.205 | Balanced −82% Summarise −86% Chat −81% Code gen −79% Agentic −84% | no extended reasoning |
| Qwen3 Next 80B A3B Instruct Qwen | 1418.6 +45.9 | $0.350 | Balanced −70% Summarise −82% Chat −65% Code gen −61% Agentic −76% | no extended reasoning |
| Qwen3 VL 235B A22B Instruct Qwen | 1420.9 +48.2 | $0.632 | Balanced −45% Summarise −67% Chat −40% Code gen −32% Agentic −59% | no extended reasoning |
| Qwen3 30B A3B Instruct 2507 Qwen | 1384.3 +11.6 | $0.084 | Balanced −93% Summarise −93% Chat −93% Code gen −92% Agentic −93% | no extended reasoning |
| DeepSeek V3 0324 DeepSeek | 1375 tie | $0.438 | Balanced −62% Summarise −64% Chat −61% Code gen −61% Agentic −63% | no extended reasoning |

## 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/).