---
title: "Compare alternatives to Qwen3.5-35B-A3B · Undominated.ai"
canonical: https://undominated.ai/alternatives/qwen__qwen3.5-35b-a3b/
description: "15 alternatives score at least as high as Qwen3.5-35B-A3B and cost no more under at least one listed workload. 5 preserve recorded capabilities; 10 have named trade-offs."
---

# Compare alternatives to Qwen3.5-35B-A3B · Undominated.ai

> 15 alternatives score at least as high as Qwen3.5-35B-A3B and cost no more under at least one listed workload. 5 preserve recorded capabilities; 10 have named trade-offs.

# Alternatives to Qwen3.5-35B-A3B

Qwen · LMArena 1394.4 · $0.355/M on a balanced workload · prices as of 2026-10-06

 [Compare models](https://undominated.ai/compare/) Comparison context

Choose up to three standard models in Compare.

Requirements and prompt length apply in Compare. This page keeps its stated price basis and workload controls.

## 5 models are higher-scoring or cheaper, with neither dimension worse.

Of the 144 models carrying both an independent score and a published price at the same delivery mode, **15** score at least as high as Qwen3.5-35B-A3B and cost no more under at least one listed workload, with at least one of those dimensions improved. **5** match or beat its recorded context, maximum output, input modes, tool use and extended reasoning. **10** have a recorded capability loss, named on every row.

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 only while a candidate improves score or price without worsening the other under a listed workload. When that comparison no longer holds, it is not generated.

## Where it holds

| Workload | Qwen3.5-35B-A3B $/M | Recorded capabilities preserved | Recorded capability losses |
| --- | --- | --- | --- |
| Balanced | $0.355 | 5 | 9 |
| Summarise | $0.157 | 5 | 7 |
| Chat | $0.457 | 5 | 10 |
| Code gen | $0.649 | 5 | 10 |
| Agentic | $0.215 | 5 | 8 |

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 | 1394.4 |
| --- | --- |
| Context window | 262K |
| Max output | 32K |
| Input modes | text, image, video |
| Tool use | yes |
| Extended reasoning | yes |

Preserving recorded capabilities requires matching or beating these fields. [Full record for Qwen3.5-35B-A3B](/models/qwen__qwen3.5-35b-a3b/), or [check Qwen3.5-35B-A3B against the whole catalogue](/check/qwen__qwen3.5-35b-a3b/).

## Recorded capabilities preserved 5

Each option has a higher score or a lower price, with neither dimension worse under the listed workloads. It matches or beats Qwen3.5-35B-A3B on recorded context, maximum output, input modes, tool use and extended reasoning.

| # | Model | Intelligence | $/M | Holds under |
| --- | --- | --- | --- | --- |
| 1 | [GLM 5.3 Flash](/models/z-ai__glm-5.3-flash/) Z.ai | 1469.6 +75.2 | $0.238 | Balanced −33% Summarise −15% Chat −44% Code gen −46% Agentic −39% |
| 2 | [MiMo-V2.6-Flash](/models/xiaomi__mimo-v2.6-flash/) Xiaomi | 1456.4 +62 | $0.175 | Balanced −51% Summarise −31% Chat −66% Code gen −67% Agentic −63% |
| 3 | [Gemma 4 26B A4B](/models/google__gemma-4-26b-a4b-it/) Google | 1433.9 +39.5 | $0.127 | Balanced −64% Summarise −56% Chat −67% Code gen −68% Agentic −63% |
| 4 | [MiMo-V2.5](/models/xiaomi__mimo-v2.5/) Xiaomi | 1427.5 +33.1 | $0.175 | Balanced −51% Summarise −31% Chat −66% Code gen −67% Agentic −63% |
| 5 | [Qwen3.5-Flash](/models/qwen__qwen3.5-flash-02-23/) Qwen | 1397 tie | $0.114 | Balanced −68% Summarise −52% Chat −69% Code gen −72% Agentic −56% |

## Alternatives with recorded capability losses 10

These improve score or price without worsening the other under the listed workloads, but reduce at least one recorded capability. Read the last column before switching.

| Model | Intelligence | $/M | Holds under | Recorded capability loss |
| --- | --- | --- | --- | --- |
| [DeepSeek V4.1 Flash](/models/deepseek__deepseek-v4.1-flash/) DeepSeek | 1462.5 +68.1 | $0.182 | Balanced −49% Summarise −58% Chat −46% Code gen −45% Agentic −51% | no video input |
| [Gemma 4 31B](/models/google__gemma-4-31b-it/) Google | 1443.4 +49 | $0.205 | Balanced −42% Summarise −3% Chat −47% Code gen −54% Agentic −17% | 33K → 16K max output |
| [Hy3](/models/tencent__hy3/) Tencent | 1440.7 +46.3 | $0.231 | Balanced −35% Summarise −21% Chat −43% Code gen −44% Agentic −39% | no image, video input |
| [DeepSeek V4 Flash 0423](/models/deepseek__deepseek-v4-flash/) DeepSeek | 1432.1 +37.7 | $0.113 | Balanced −68% Summarise −53% Chat −77% Code gen −79% Agentic −72% | no image, video input |
| [DeepSeek V3.2](/models/deepseek__deepseek-v3.2/) DeepSeek | 1424.5 +30.1 | $0.290 | Balanced −18% Chat −41% Code gen −50% Agentic −7% | 262K → 164K context no image, video input |
| [DeepSeek V3.2 Exp](/models/deepseek__deepseek-v3.2-exp/) DeepSeek | 1422.4 +28 | $0.305 | Balanced −14% Chat −29% Code gen −45% | 262K → 164K context no image, video input |
| [Qwen3 235B A22B Instruct 2507](/models/qwen__qwen3-235b-a22b-2507/) Qwen | 1419.3 +24.9 | $0.205 | Balanced −42% Summarise −28% Chat −40% Code gen −44% Agentic −26% | 33K → 16K max output no image, video input no extended reasoning |
| [Qwen3 Next 80B A3B Instruct](/models/qwen__qwen3-next-80b-a3b-instruct/) Qwen | 1417.3 +22.9 | $0.268 | Balanced −24% Summarise −16% Chat −19% Code gen −22% Agentic −7% | no image, video input no extended reasoning |
| [Step 3.5 Flash](/models/stepfun__step-3.5-flash/) StepFun | 1403 tie | $0.150 | Balanced −58% Summarise −30% Chat −61% Code gen −66% Agentic −40% | no image, video input |
| [Qwen3 VL 235B A22B Instruct](/models/qwen__qwen3-vl-235b-a22b-instruct/) Qwen | 1420.1 +25.7 | $0.445 chat | Chat −3% Code gen −7% | no video input 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 ±10.47 points, so a gap smaller than that is marked *tie* rather than an improvement — see [significance bands](/significance/).
 - 204 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.
 - Models with no higher-scoring or cheaper alternative that is no worse on the other axis are on the [value frontier](/frontier/). Models with such an alternative are [listed here](/alternatives/).

## Continue your investigation

 - [Check requirements](/check/)
- [Compare exact differences](/compare/)
- [Set a quality floor](/cheapest-at/)
- [Inspect tier boundaries](/cliffs/)