# GPT-5.2-Codex vs GPT-5.4 mini

> Side-by-side benchmark comparison for GPT-5.2-Codex and GPT-5.4 mini across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/gpt-5-2-codex-vs-gpt-5-4-mini
- Last updated: September 10, 2026

- Shared sourced benchmarks: 1
- HTML indexing: indexable
- Ranking lane: BenchAlign v5

## Quick Verdict

Evidence is too thin for a benchmark verdict: GPT-5.2-Codex and GPT-5.4 mini share only 1 sourced benchmark row.

## Summary

- Only 1 shared sourced benchmark row is available. That is not enough for a broad capability verdict, so this mirror does not name an overall winner.
- GPT-5.4 mini has the lower published output-token price.

## Model Snapshot

| Property | GPT-5.2-Codex | GPT-5.4 mini |
|----------|----------|----------|
| Creator | OpenAI | OpenAI |
| Type | Proprietary | Proprietary |
| Reasoning | Reasoning | Reasoning |
| Context | 400K | 400K |
| Independent public score (not head-to-head) | 54.19 | 61.11 |
| Benchmarks Covered | 3 | 19 |
| Pricing (input/output) | $1.75 / $14.00 | $0.75 / $4.50 |

## Category Breakdown

### Agentic

- Winner: Insufficient shared evidence
- GPT-5.2-Codex public-lane score: 50.6 (Estimated · #47/152)
- GPT-5.4 mini public-lane score: 39.1 (Estimated · #119/152)

| Benchmark | GPT-5.2-Codex | GPT-5.4 mini | Winner |
|-----------|-----------|-----------|--------|
| Gert Labs | 51.79% | Coming soon | Coming soon |
| JobBench | 26.0% | Coming soon | Coming soon |
| Terminal-Bench 2.0 | Coming soon | 60% | Coming soon |
| OSWorld-Verified | Coming soon | 72.1% | Coming soon |
| MCP Atlas | Coming soon | 57.7% | Coming soon |
| Toolathlon | Coming soon | 42.9% | Coming soon |
| τ²-bench results | Coming soon | 93.4% | Coming soon |
| Terminal-Bench 2.1 (Vals) | Coming soon | 54.7% | Coming soon |

### Coding

- Winner: Insufficient shared evidence
- GPT-5.2-Codex public-lane score: 51.5 (Supported · #50/151)
- GPT-5.4 mini public-lane score: 42.7 (Supported · #104/151)

| Benchmark | GPT-5.2-Codex | GPT-5.4 mini | Winner |
|-----------|-----------|-----------|--------|
| Vibe Code Bench | 37.91% | 47.97% | GPT-5.4 mini |
| FrontierCode 1.1 Main | Coming soon | 27.0% | Coming soon |
| LiveCodeBench (Vals) | Coming soon | 81.5% | Coming soon |
| SWE-bench (Vals) | Coming soon | 73.0% | Coming soon |

### Multimodal & Grounded

- Winner: Insufficient shared evidence
- GPT-5.2-Codex public-lane score: 72.4 (Unranked · 1 rankable row)
- GPT-5.4 mini public-lane score: 57.2 (#31/48)

| Benchmark | GPT-5.2-Codex | GPT-5.4 mini | Winner |
|-----------|-----------|-----------|--------|
| MMMU-Pro | Coming soon | 76.6% | Coming soon |
| MMMU-Pro w/ Python | Coming soon | 78% | Coming soon |

### Reasoning

- Winner: Insufficient shared evidence
- GPT-5.2-Codex public-lane score: 77.6 (Unranked · 2 rankable rows)
- GPT-5.4 mini public-lane score: 73.9 (Unranked · 2 rankable rows)

### Knowledge

- Winner: Insufficient shared evidence
- GPT-5.2-Codex public-lane score: 56.2 (Estimated · #50/183)
- GPT-5.4 mini public-lane score: 55.6 (Supported · #52/183)

| Benchmark | GPT-5.2-Codex | GPT-5.4 mini | Winner |
|-----------|-----------|-----------|--------|
| GPQA | Coming soon | 88% | Coming soon |
| HLE | Coming soon | 41.5% | Coming soon |
| HLE w/o tools | Coming soon | 28.2% | Coming soon |
| GPQA Diamond (Vals) | Coming soon | 83.1% | Coming soon |
| MMLU-Pro (Vals) | Coming soon | 84.6% | Coming soon |

### Instruction Following

- Winner: Insufficient shared evidence
- GPT-5.2-Codex public-lane score: 93.7 (#3/123)
- GPT-5.4 mini public-lane score: 89.8 (#23/123)

### Mathematics

- Winner: Insufficient shared evidence
- GPT-5.2-Codex public-lane score: Coming soon
- GPT-5.4 mini public-lane score: 44.5 (Unranked · 2 rankable rows)

| Benchmark | GPT-5.2-Codex | GPT-5.4 mini | Winner |
|-----------|-----------|-----------|--------|
| FrontierMath v2 (Tiers 1-3) | Coming soon | 28.280% | Coming soon |
| FrontierMath v2 (Tier 4) | Coming soon | 2.080% | Coming soon |

## FAQ

### Is there enough evidence to pick between GPT-5.2-Codex and GPT-5.4 mini?

Not yet. The models share only 1 sourced benchmark row, so this mirror reports the available measurements and metadata without naming a broad winner.

### What can I compare safely on this page?

Published pricing, context windows, model type, and individual shared benchmark rows are directly comparable. Independent overall scores and non-overlapping benchmark coverage are labeled separately and should not be read as a controlled head-to-head result.

## Related Comparisons

- [GPT-5.2-Codex vs GPT-5.4](/compare/gpt-5-2-codex-vs-gpt-5-4)
- [GPT-5.4 mini vs GPT-5.4](/compare/gpt-5-4-vs-gpt-5-4-mini)
- [GPT-5.2-Codex vs GPT-5.4 Pro](/compare/gpt-5-2-codex-vs-gpt-5-4-pro)
- [GPT-5.4 mini vs GPT-5.4 Pro](/compare/gpt-5-4-mini-vs-gpt-5-4-pro)

## Explore More

- [GPT-5.2-Codex profile](/models/gpt-5-2-codex)
- [GPT-5.4 mini profile](/models/gpt-5-4-mini)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)