---
title: "Cheaper alternatives to Qwen3 235B A22B Instruct 2507 · Undominated.ai"
canonical: https://undominated.ai/alternatives/qwen__qwen3-235b-a22b-2507/
description: "7 models score at least as high as Qwen3 235B A22B Instruct 2507 and cost no more on a balanced workload. 5 give up nothing measurable; the other 2 name what they drop. Measured prices and independent scores."
---

# Cheaper alternatives to Qwen3 235B A22B Instruct 2507 · Undominated.ai

> 7 models score at least as high as Qwen3 235B A22B Instruct 2507 and cost no more on a balanced workload. 5 give up nothing measurable; the other 2 name what they drop. Measured prices and independent scores.

# Cheaper alternatives to Qwen3 235B A22B Instruct 2507

Qwen · LMArena 1419.3 · $0.205/M on a balanced workload · prices as of 2026-08-28

## 5 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, **7** score at least as high as Qwen3 235B A22B Instruct 2507 and cost no more under at least one workload. **5** match or beat it on every capability we hold — context, maximum output, input modes, tool use and extended reasoning. **2** 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 Qwen3 235B A22B Instruct 2507, it is not generated.

## Where it holds

| Workload | Qwen3 235B A22B Instruct 2507 $/M | Better and cheaper | With a trade |
| --- | --- | --- | --- |
| Balanced | $0.205 | 5 | 0 |
| Summarise | $0.113 | 5 | 0 |
| Chat | $0.274 | 5 | 1 |
| Code gen | $0.366 | 5 | 2 |
| Agentic | $0.159 | 5 | 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 | 1419.3 |
| --- | --- |
| Context window | 262K |
| Max output | 16K |
| Input modes | text |
| Tool use | yes |
| Extended reasoning | no |

A replacement has to clear every line above, not just the score. [Full record for Qwen3 235B A22B Instruct 2507](/models/qwen__qwen3-235b-a22b-2507/).

## Better and cheaper, nothing given up 5

Each of these matches or beats Qwen3 235B A22B Instruct 2507 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 | Gemma 4 31B Google | 1441.7 +22.4 | $0.152 | Balanced −26% Summarise −19% Chat −35% Code gen −35% Agentic −35% |
| 2 | Hy3 Tencent | 1441.2 +21.9 | $0.144 | Balanced −30% Summarise −32% Chat −41% Code gen −38% Agentic −48% |
| 3 | Gemma 4 26B A4B Google | 1434.6 +15.3 | $0.138 | Balanced −33% Summarise −26% Chat −35% Code gen −37% Agentic −31% |
| 4 | DeepSeek V4 Flash 0423 DeepSeek | 1431.6 +12.3 | $0.101 | Balanced −51% Summarise −41% Chat −66% Code gen −66% Agentic −66% |
| 5 | MiMo-V2.5 Xiaomi | 1427.3 tie | $0.175 | Balanced −15% Summarise −5% Chat −44% Code gen −42% Agentic −50% |

## Cheaper and higher-scoring, but you give something up 2

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 |
| --- | --- | --- | --- | --- |
| DeepSeek V3.2 DeepSeek | 1424.6 tie | $0.269 chat | Chat −2% Code gen −12% | 262K → 164K context |
| DeepSeek V3.2 Exp DeepSeek | 1424.4 tie | $0.354 code gen | Code gen −3% | 262K → 164K context |

## 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/).