# GPT-5.3 Codex vs MAI-Thinking-1

> Side-by-side benchmark comparison for GPT-5.3 Codex and MAI-Thinking-1 across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/gpt-5-3-codex-vs-mai-thinking-1
- Last updated: September 4, 2026

- Shared sourced benchmarks: 3
- HTML indexing: indexable
- Ranking lane: BenchAlign v5

## Quick Verdict

Evidence is too thin for a benchmark verdict: GPT-5.3 Codex and MAI-Thinking-1 share only 3 sourced benchmark rows.

## Summary

- Only 3 shared sourced benchmark rows are available. That is not enough for a broad capability verdict, so this mirror does not name an overall winner.
- GPT-5.3 Codex has the larger published context window at 400K.

## Model Snapshot

| Property | GPT-5.3 Codex | MAI-Thinking-1 |
|----------|----------|----------|
| Creator | OpenAI | Microsoft |
| Type | Proprietary | Proprietary |
| Reasoning | Reasoning | Reasoning |
| Context | 400K | 256K |
| Independent public score (not head-to-head) | 65.22 | 52.27 |
| Benchmarks Covered | 8 | 14 |
| Pricing (input/output) | $1.75 / $14.00 | N/A |

## Category Breakdown

### Agentic

- Winner: Insufficient shared evidence
- GPT-5.3 Codex public-lane score: 62.3 (Estimated · #19/151)
- MAI-Thinking-1 public-lane score: 51.7 (Estimated · #53/151)

| Benchmark | GPT-5.3 Codex | MAI-Thinking-1 | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 2.0 | 77.3% | 46% | GPT-5.3 Codex |
| OSWorld-Verified | 64.7% | Coming soon | Coming soon |
| Gert Labs | 57.47% | Coming soon | Coming soon |
| JobBench | 33.7% | Coming soon | Coming soon |

### Coding

- Winner: Insufficient shared evidence
- GPT-5.3 Codex public-lane score: 62.7 (Supported · #16/183)
- MAI-Thinking-1 public-lane score: 51.8 (Estimated · #60/183)

| Benchmark | GPT-5.3 Codex | MAI-Thinking-1 | Winner |
|-----------|-----------|-----------|--------|
| SWE-bench Verified | 85% | 73.5% | GPT-5.3 Codex |
| SWE-bench Pro | 56.8% | 52.8% | GPT-5.3 Codex |
| SWE-Rebench | 58.2% | Coming soon | Coming soon |
| Vibe Code Bench | 61.77% | Coming soon | Coming soon |
| LiveCodeBench v6 | Coming soon | 87.7% | Coming soon |
| Terminal-Bench 2.0 | Coming soon | 46.0% | Coming soon |

### Multimodal & Grounded

- Winner: Insufficient shared evidence
- GPT-5.3 Codex public-lane score: 75.6 (Unranked · 1 rankable row)
- MAI-Thinking-1 public-lane score: Coming soon

### Reasoning

- Winner: Insufficient shared evidence
- GPT-5.3 Codex public-lane score: 77.3 (Unranked · 2 rankable rows)
- MAI-Thinking-1 public-lane score: Coming soon

| Benchmark | GPT-5.3 Codex | MAI-Thinking-1 | Winner |
|-----------|-----------|-----------|--------|
| Graphwalks BFS 128K | Coming soon | 90% | Coming soon |

### Knowledge

- Winner: Insufficient shared evidence
- GPT-5.3 Codex public-lane score: 66.3 (Estimated · #22/181)
- MAI-Thinking-1 public-lane score: 53.3 (Estimated · #68/181)

| Benchmark | GPT-5.3 Codex | MAI-Thinking-1 | Winner |
|-----------|-----------|-----------|--------|
| GPQA | Coming soon | 84.2% | Coming soon |
| GPQA-D | Coming soon | 84.2% | Coming soon |
| MMLU-Pro | Coming soon | 85% | Coming soon |
| SimpleQA | Coming soon | 31% | Coming soon |

### Instruction Following

- Winner: Insufficient shared evidence
- GPT-5.3 Codex public-lane score: 92.3 (#15/120)
- MAI-Thinking-1 public-lane score: 94.7 (#1/120)

| Benchmark | GPT-5.3 Codex | MAI-Thinking-1 | Winner |
|-----------|-----------|-----------|--------|
| IFBench | Coming soon | 85% | Coming soon |

### Mathematics

- Winner: Insufficient shared evidence
- GPT-5.3 Codex public-lane score: Coming soon
- MAI-Thinking-1 public-lane score: 73.4 (Unranked · 3 rankable rows)

| Benchmark | GPT-5.3 Codex | MAI-Thinking-1 | Winner |
|-----------|-----------|-----------|--------|
| AIME 2025 | Coming soon | 97% | Coming soon |
| AIME26 | Coming soon | 94.5% | Coming soon |
| HMMT Feb 2026 | Coming soon | 84.9% | Coming soon |

## FAQ

### Is there enough evidence to pick between GPT-5.3 Codex and MAI-Thinking-1?

Not yet. The models share only 3 sourced benchmark rows, 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-5.3 Codex vs GPT-5.3-Codex-Spark](/compare/gpt-5-3-codex-vs-gpt-5-3-codex-spark)
- [MAI-Thinking-1 vs GPT-5.3-Codex-Spark](/compare/gpt-5-3-codex-spark-vs-mai-thinking-1)
- [GPT-5.3 Codex vs Claude Fable 5.1](/compare/claude-fable-5-1-vs-gpt-5-3-codex)
- [MAI-Thinking-1 vs Claude Fable 5.1](/compare/claude-fable-5-1-vs-mai-thinking-1)

## Explore More

- [GPT-5.3 Codex profile](/models/gpt-5-3-codex)
- [MAI-Thinking-1 profile](/models/mai-thinking-1)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)