Model Comparison

GPT-5.4 vs Claude Sonnet 5Which is better in 2026?

Claude Sonnet 5 significantly outperforms across most benchmarks. GPT-5.4 is 1.1x cheaper per token.

Verdict: GPT-5.4 vs Claude Sonnet 5 — which is better?

GPT-5.4 (by OpenAI) and Claude Sonnet 5 (by Anthropic) are two of the AI models people compare most. Here is how they stack up on benchmarks, price and capabilities, and which one to pick in 2026.

GPT-5.4 outperforms in 1 benchmarks (Toolathlon), while Claude Sonnet 5 is better at 7 benchmarks (BrowseComp, GDPval-AA, Humanity's Last Exam, Legal Agent Benchmark, OSWorld-Verified, SWE-Bench Pro, Terminal-Bench 2.0). Claude Sonnet 5 significantly outperforms across most benchmarks.

On price, GPT-5.4 is roughly 1.1x cheaper per token on a blended 3:1 input/output basis, which adds up quickly at production volume.

Choose GPT-5.4 if…

  • cost matters — it's about 1.1x cheaper per token

Choose Claude Sonnet 5 if…

  • you want the strongest raw capability — it leads on 7 of 8 shared benchmarks
  • you want the most recent training data — it shipped Jun 2026

Performance Benchmarks

Comparative analysis across standard metrics

8 benchmarks

GPT-5.4 outperforms in 1 benchmarks (Toolathlon), while Claude Sonnet 5 is better at 7 benchmarks (BrowseComp, GDPval-AA, Humanity's Last Exam, Legal Agent Benchmark, OSWorld-Verified, SWE-Bench Pro, Terminal-Bench 2.0).

Claude Sonnet 5 significantly outperforms across most benchmarks.

Tue Jun 30 2026 • llm-stats.com

Arena Performance

Human preference votes

Pricing Analysis

Price comparison per million tokens

GPT-5.4 costs less

For input processing, GPT-5.4 ($2.50/1M tokens) is 1.2x cheaper than Claude Sonnet 5 ($3.00/1M tokens).

For output processing, GPT-5.4 ($15.00/1M tokens) costs the same as Claude Sonnet 5 ($15.00/1M tokens).

In conclusion, Claude Sonnet 5 is more expensive than GPT-5.4.*

* Using a 3:1 ratio of input to output tokens

Lowest available price from all providers
Tue Jun 30 2026 • llm-stats.com
OpenAI
GPT-5.4
Input tokens$2.50
Output tokens$15.00
Best providerOpenAI
Anthropic
Claude Sonnet 5
Input tokens$3.00
Output tokens$15.00
Best providerAnthropic
Notice missing or incorrect data?Start an Issue

Context Window

Maximum input and output token capacity

Both models have the same input context window of 1,000,000 tokens. GPT-5.4 can generate longer responses up to 128,000 tokens, while Claude Sonnet 5 is limited to 64,000 tokens.

OpenAI
GPT-5.4
Input1,000,000 tokens
Output128,000 tokens
Anthropic
Claude Sonnet 5
Input1,000,000 tokens
Output64,000 tokens
Tue Jun 30 2026 • llm-stats.com

Input Capabilities

Supported data types and modalities

Both GPT-5.4 and Claude Sonnet 5 support multimodal inputs.

They are both capable of processing various types of data, offering versatility in application.

GPT-5.4

Text
Images
Audio
Video

Claude Sonnet 5

Text
Images
Audio
Video

License

Usage and distribution terms

Both models are licensed under proprietary licenses.

Both models have usage restrictions defined by their respective organizations.

GPT-5.4

Proprietary

Closed source

Claude Sonnet 5

Proprietary

Closed source

Release Timeline

When each model was launched

GPT-5.4 was released on 2026-03-05, while Claude Sonnet 5 was released on 2026-06-30.

Claude Sonnet 5 is 4 months newer than GPT-5.4.

GPT-5.4

Mar 5, 2026

3 months ago

Claude Sonnet 5

Jun 30, 2026

0 days ago

3mo newer

Knowledge Cutoff

When training data ends

Neither model specifies a knowledge cutoff date.

Unable to compare the recency of their training data.

No cutoff dates available

Provider Availability

GPT-5.4 is available from OpenAI. Claude Sonnet 5 is available from Anthropic.

GPT-5.4

openai logo
OpenAI
Input Price:Input: $2.50/1MOutput Price:Output: $15.00/1M

Claude Sonnet 5

anthropic logo
Anthropic
Input Price:Input: $3.00/1MOutput Price:Output: $15.00/1M
* Prices shown are per million tokens

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion

Key Takeaways

Less expensive input tokens
Higher Toolathlon score (54.6% vs 54.3%)
Higher BrowseComp score (84.7% vs 82.7%)
Higher GDPval-AA score (53.9% vs 47.6%)
Higher Humanity's Last Exam score (57.4% vs 39.8%)
Higher Legal Agent Benchmark score (5.8% vs 0.4%)
Higher OSWorld-Verified score (81.2% vs 75.0%)
Higher SWE-Bench Pro score (63.2% vs 57.7%)
Higher Terminal-Bench 2.0 score (80.4% vs 75.1%)

Detailed Comparison

Interactive Arena

Judge for yourself.

Run your own prompts against GPT-5.4 and Claude Sonnet 5 side-by-side, then vote on the output you prefer.

GPT-5.4
✓ Preferred
Claude Sonnet 5
Open in Playground
AI Model Comparison Table
Feature
OpenAI
GPT-5.4
Anthropic
Claude Sonnet 5

FAQ

Common questions about GPT-5.4 vs Claude Sonnet 5.

Which is better, GPT-5.4 or Claude Sonnet 5?

Claude Sonnet 5 significantly outperforms across most benchmarks. GPT-5.4 is made by OpenAI and Claude Sonnet 5 is made by Anthropic. The best choice depends on your use case — compare their benchmark scores, pricing, and capabilities above.

How does GPT-5.4 compare to Claude Sonnet 5 in benchmarks?

GPT-5.4 scores Tau2 Telecom: 98.9%, ARC-AGI: 93.7%, Graphwalks BFS <128k: 93.0%, GPQA: 92.8%, Graphwalks parents <128k: 89.8%. Claude Sonnet 5 scores CharXiv-R: 88.3%, ChartMuseum: 86.7%, SWE-Bench Verified: 85.2%, BrowseComp: 84.7%, GDP.pdf: 81.6%.

Is GPT-5.4 cheaper than Claude Sonnet 5?

GPT-5.4 is 1.2x cheaper for input tokens. GPT-5.4 costs $2.50/M input and $15.00/M output via openai. Claude Sonnet 5 costs $3.00/M input and $15.00/M output via anthropic.

What are the context window sizes for GPT-5.4 and Claude Sonnet 5?

GPT-5.4 supports 1.0M tokens and Claude Sonnet 5 supports 1.0M tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between GPT-5.4 and Claude Sonnet 5?

Key differences include input pricing ($2.50 vs $3.00/M). See the full comparison above for benchmark-by-benchmark results.

Who makes GPT-5.4 and Claude Sonnet 5?

GPT-5.4 is developed by OpenAI and Claude Sonnet 5 is developed by Anthropic.