---
title: "Cheaper alternatives to GPT-5.4 · Undominated.ai"
canonical: https://undominated.ai/alternatives/openai__gpt-5.4/
description: "7 models score at least as high as GPT-5.4 and cost no more on a balanced workload. 0 give up nothing measurable; the other 7 name what they drop. Measured prices and independent scores."
---

# Cheaper alternatives to GPT-5.4 · Undominated.ai

> 7 models score at least as high as GPT-5.4 and cost no more on a balanced workload. 0 give up nothing measurable; the other 7 name what they drop. Measured prices and independent scores.

# Cheaper alternatives to GPT-5.4

OpenAI · LMArena 1469.6 · $5.63/M on a balanced workload · prices as of 2026-08-28

## Everything cheaper and higher-scoring gives something up.

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 GPT-5.4 and cost no more under at least one workload. **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 GPT-5.4, it is not generated.

## Where it holds

| Workload | GPT-5.4 $/M | Better and cheaper | With a trade |
| --- | --- | --- | --- |
| Balanced | $5.63 | 0 | 7 |
| Summarise | $2.48 | 0 | 7 |
| Chat | $6.83 | 0 | 7 |
| Code gen | $9.82 | 0 | 7 |
| Agentic | $3.04 | 0 | 7 |

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 | 1469.6 |
| --- | --- |
| Context window | 1.1M |
| Max output | 128K |
| Input modes | text, image, file |
| Tool use | yes |
| Extended reasoning | yes |

A replacement has to clear every line above, not just the score. [Full record for GPT-5.4](/models/openai__gpt-5.4/).

## 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 |
| --- | --- | --- | --- | --- |
| Gemini 3.7 Flash Google | 1490.2 +20.6 | $0.750 | Balanced −87% Summarise −86% Chat −87% Code gen −87% Agentic −87% | 1.1M → 1.0M context 128K → 66K max output |
| Gemini 3.5 Flash Google | 1482.6 +13 | $3.38 | Balanced −40% Summarise −40% Chat −40% Code gen −40% Agentic −40% | 1.1M → 1.0M context 128K → 66K max output |
| Qwen3.8 Max Qwen | 1481.9 +12.3 | $3.00 | Balanced −47% Summarise −32% Chat −55% Code gen −57% Agentic −49% | 1.1M → 1.0M context no file input |
| Muse Spark 1.1 Meta | 1478.3 tie | $2.00 | Balanced −64% Summarise −56% Chat −69% Code gen −70% Agentic −66% | 1.1M → 1.0M context |
| Gemini 3.6 Flash Google | 1476.5 tie | $1.50 | Balanced −73% Summarise −72% Chat −74% Code gen −75% Agentic −74% | 1.1M → 1.0M context 128K → 66K max output |
| GLM 5.3 Z.ai | 1476.9 tie | $2.15 | Balanced −62% Summarise −51% Chat −67% Code gen −68% Agentic −61% | 1.1M → 1.0M context no image, file input |
| Gemini 3.1 Pro Preview Google | 1479.5 tie | $4.50 | Balanced −20% Summarise −20% Chat −20% Code gen −20% Agentic −20% | 1.1M → 1.0M context 128K → 66K 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/).