# Claude 4 Sonnet vs GPT-5.2

> Side-by-side benchmark comparison for Claude 4 Sonnet and GPT-5.2 across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/claude-4-sonnet-vs-gpt-5-2
- Last updated: October 17, 2026

- Shared sourced benchmarks: 4
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.8

## Quick Verdict

GPT-5.2 leads by point estimate. Conditional score ranges do not establish rank confidence. Choose using the category evidence and your own workload tests.

## Summary

- GPT-5.2 has the higher overall point estimate 59.86 to 38.08. Conditional score ranges do not establish rank confidence.
- The clearest category separation is in instruction following, where the averages are 53.5 for Claude 4 Sonnet and 92.6 for GPT-5.2.
- The biggest single benchmark swing is AIME 2025 in Mathematics, with scores of 33.1% and 100%.
- GPT-5.2 is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.
- GPT-5.2 also has the larger context window at 400K.

## Model Snapshot

| Property | Claude 4 Sonnet | GPT-5.2 |
|----------|----------|----------|
| Creator | Anthropic | OpenAI |
| Type | Proprietary | Proprietary |
| Reasoning | Non-Reasoning | Reasoning |
| Context | 200K | 400K |
| Overall Score | 38.08 | 59.86 |
| Benchmarks Covered | 4 | 31 |
| Pricing (input/output) | $3.00 / $15.00 | $1.75 / $14.00 |

## Category Breakdown

### Agentic

- Winner: Not comparable
- Claude 4 Sonnet public-lane score: Coming soon
- GPT-5.2 public-lane score: 44.4 (Supported · #58/128)

| Benchmark | Claude 4 Sonnet | GPT-5.2 | Winner |
|-----------|-----------|-----------|--------|
| Gert Labs | 39.66% | 46.54% | GPT-5.2 |
| JobBench | 18.4% | 34.3% | GPT-5.2 |
| BrowseComp | Coming soon | 65.8% | Coming soon |
| OSWorld-Verified | Coming soon | 47.3% | Coming soon |

### Coding

- Winner: Not comparable
- Claude 4 Sonnet public-lane score: Coming soon
- GPT-5.2 public-lane score: 36.8 (Supported · #73/150)

| Benchmark | Claude 4 Sonnet | GPT-5.2 | Winner |
|-----------|-----------|-----------|--------|
| SWE-bench Verified | 72.7% | 80% | GPT-5.2 |
| SWE-bench Pro | Coming soon | 55.6% | Coming soon |
| Vibe Code Bench | Coming soon | 53.50% | Coming soon |

### Reasoning

- Winner: Not comparable
- Claude 4 Sonnet public-lane score: 56.1 (Unranked · 2 rankable rows)
- GPT-5.2 public-lane score: 64.3 (Unranked · 4 rankable rows)

| Benchmark | Claude 4 Sonnet | GPT-5.2 | Winner |
|-----------|-----------|-----------|--------|
| ARC-AGI-2 | Coming soon | 52.9% | Coming soon |
| ARC-AGI-1 | Coming soon | 86.17% | Coming soon |

### Multimodal & Grounded

- Winner: Not comparable
- Claude 4 Sonnet public-lane score: 56.4 (Unranked · 1 rankable row)
- GPT-5.2 public-lane score: 75.9 (#22/54)

| Benchmark | Claude 4 Sonnet | GPT-5.2 | Winner |
|-----------|-----------|-----------|--------|
| MMMU | Coming soon | 86.7% | Coming soon |
| MMMU-Pro | Coming soon | 79.5% | Coming soon |
| RealWorldQA | Coming soon | 83.3% | Coming soon |
| OmniDocBench 1.5 | Coming soon | 85.7% | Coming soon |
| Video-MME (with subtitle) | Coming soon | 86.0% | Coming soon |
| Video-MME (w/o subtitle) | Coming soon | 85.8% | Coming soon |
| MathVision | Coming soon | 83.0% | Coming soon |
| We-Math | Coming soon | 79.0% | Coming soon |
| DynaMath | Coming soon | 86.8% | Coming soon |
| SimpleVQA | Coming soon | 55.8% | Coming soon |
| CC-OCR | Coming soon | 70.3% | Coming soon |
| AI2D_TEST | Coming soon | 92.2% | Coming soon |
| CountBench | Coming soon | 91.9% | Coming soon |
| ERQA | Coming soon | 59.8% | Coming soon |
| VideoMMMU | Coming soon | 85.9% | Coming soon |
| MLVU (M-Avg) | Coming soon | 85.6% | Coming soon |
| CharXiv | Coming soon | 82.1% | Coming soon |
| V* | Coming soon | 75.9% | Coming soon |

### Knowledge

- Winner: Directional only
- Claude 4 Sonnet public-lane score: 37.7 (Estimated · #124/184)
- GPT-5.2 public-lane score: 59.7 (Supported · #50/184)

| Benchmark | Claude 4 Sonnet | GPT-5.2 | Winner |
|-----------|-----------|-----------|--------|
| GPQA | Coming soon | 92.4% | Coming soon |

### Instruction Following

- Winner: Directional only
- Claude 4 Sonnet public-lane score: 53.5 (#86/148)
- GPT-5.2 public-lane score: 92.6 (#19/148)

### Mathematics

- Winner: Not comparable
- Claude 4 Sonnet public-lane score: Coming soon
- GPT-5.2 public-lane score: 71.1 (Unranked · 3 rankable rows)

| Benchmark | Claude 4 Sonnet | GPT-5.2 | Winner |
|-----------|-----------|-----------|--------|
| AIME 2025 | 33.1% | 100% | GPT-5.2 |
| FrontierMath v2 (Tiers 1-3) | Coming soon | 67.400% | Coming soon |
| FrontierMath v2 (Tier 4) | Coming soon | 31.700% | Coming soon |

## FAQ

### Which is better overall, Claude 4 Sonnet or GPT-5.2?

GPT-5.2 has the higher overall point estimate. Conditional score ranges do not establish rank confidence.

### Where is the biggest gap between Claude 4 Sonnet and GPT-5.2?

The widest category gap is in instruction following, where the averages are 53.5 for Claude 4 Sonnet and 92.6 for GPT-5.2.

### How many benchmarks does Claude 4 Sonnet cover on BenchLM?

Claude 4 Sonnet currently has 4 sourced benchmark scores on BenchLM.

### How many benchmarks does GPT-5.2 cover on BenchLM?

GPT-5.2 currently has 31 sourced benchmark scores on BenchLM.

## Related Comparisons

- [Claude 4 Sonnet vs GPT-5.2 Pro](/compare/claude-4-sonnet-vs-gpt-5-2-pro)
- [GPT-5.2 vs GPT-5.2 Pro](/compare/gpt-5-2-vs-gpt-5-2-pro)
- [Claude 4 Sonnet vs GPT-5.2 Instant](/compare/claude-4-sonnet-vs-gpt-5-2-instant)
- [GPT-5.2 vs GPT-5.2 Instant](/compare/gpt-5-2-vs-gpt-5-2-instant)

## Explore More

- [Claude 4 Sonnet profile](/models/claude-4-sonnet)
- [GPT-5.2 profile](/models/gpt-5-2)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)