---
title: "Compare alternatives to Phi 4 · Undominated.ai"
canonical: https://undominated.ai/alternatives/microsoft__phi-4/
description: "13 alternatives score at least as high as Phi 4 and cost no more under at least one listed workload. 13 preserve recorded capabilities; 0 have named trade-offs."
---

# Compare alternatives to Phi 4 · Undominated.ai

> 13 alternatives score at least as high as Phi 4 and cost no more under at least one listed workload. 13 preserve recorded capabilities; 0 have named trade-offs.

# Alternatives to Phi 4

Microsoft · LMArena 1216.7 · $0.088/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.

## 13 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, **13** score at least as high as Phi 4 and cost no more under at least one listed workload, with at least one of those dimensions improved. **13** match or beat its recorded 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 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 | Phi 4 $/M | Recorded capabilities preserved | Recorded capability losses |
| --- | --- | --- | --- |
| Balanced | $0.088 | 5 | 0 |
| Summarise | $0.074 | 10 | 0 |
| Chat | $0.098 | 5 | 0 |
| Code gen | $0.112 | 4 | 0 |
| Agentic | $0.081 | 11 | 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 | 1216.7 |
| --- | --- |
| Context window | 16K |
| Max output | 14K |
| Input modes | text |
| Tool use | no |
| Extended reasoning | no |

Preserving recorded capabilities requires matching or beating these fields. [Full record for Phi 4](/models/microsoft__phi-4/), or [check Phi 4 against the whole catalogue](/check/microsoft__phi-4/).

## Recorded capabilities preserved 13

Each option has a higher score or a lower price, with neither dimension worse under the listed workloads. It matches or beats Phi 4 on recorded context, maximum output, input modes, tool use and extended reasoning.

| # | Model | Intelligence | $/M | Holds under |
| --- | --- | --- | --- | --- |
| 1 | [gpt-oss-120b](/models/openai__gpt-oss-120b/) OpenAI | 1365.4 +148.7 | $0.065 | Balanced −26% Summarise −50% Chat −12% Agentic −37% |
| 2 | [Gemma 3 12B](/models/google__gemma-3-12b-it/) Google | 1334.2 +117.5 | $0.075 | Balanced −14% Summarise −25% Chat −8% Code gen −2% Agentic −19% |
| 3 | [Gemma 3 4B](/models/google__gemma-3-4b-it/) Google | 1290.8 +74.1 | $0.063 | Balanced −29% Summarise −29% Chat −29% Code gen −29% Agentic −29% |
| 4 | [gpt-oss-20b](/models/openai__gpt-oss-20b/) OpenAI | 1287.7 +71 | $0.036 | Balanced −59% Summarise −74% Chat −55% Code gen −46% Agentic −71% |
| 5 | [Mistral Small 3](/models/mistralai__mistral-small-24b-instruct-2501/) Mistral | 1233.6 +16.9 | $0.058 | Balanced −34% Summarise −30% Chat −37% Code gen −39% Agentic −32% |
| 6 | [DeepSeek V4.1 Flash](/models/deepseek__deepseek-v4.1-flash/) DeepSeek | 1462.5 +245.8 | $0.067 summarise | Summarise −9% |
| 7 | [MiMo-V2.6-Flash](/models/xiaomi__mimo-v2.6-flash/) Xiaomi | 1456.4 +239.7 | $0.079 agentic | Agentic −1% |
| 8 | [DeepSeek V4 Flash 0423](/models/deepseek__deepseek-v4-flash/) DeepSeek | 1432.1 +215.4 | $0.061 agentic | Agentic −25% |
| 9 | [Gemma 4 26B A4B](/models/google__gemma-4-26b-a4b-it/) Google | 1433.9 +217.2 | $0.069 summarise | Summarise −6% Agentic −1% |
| 10 | [MiMo-V2.5](/models/xiaomi__mimo-v2.5/) Xiaomi | 1427.5 +210.8 | $0.079 agentic | Agentic −1% |
| 11 | [GLM 4.7 Flash](/models/z-ai__glm-4.7-flash/) Z.ai | 1350.9 +134.2 | $0.063 summarise | Summarise −15% |
| 12 | [GPT-5 Nano](/models/openai__gpt-5-nano/) OpenAI | 1320 +103.3 | $0.055 summarise | Summarise −26% Agentic −6% |
| 13 | [Granite 4.2 8B](/models/ibm-granite__granite-4.2-8b/) IBM | 1317.8 +101.1 | $0.057 summarise | Summarise −23% Agentic −23% |

## 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/)