Price Per TokenPrice Per Token
Anthropic
Anthropic
vs
OpenAI
OpenAI

Claude Sonnet 4 vs GPT-5

A detailed comparison of pricing, benchmarks, and capabilities

Sponsor Price Per Token Reach 5000+ developers comparing LLM APIs

116 out of our 296 tracked models have had a price change in January.

Make informed model choices with updates on pricing, new releases, and tools with our weekly newsletter.

Key Takeaways

Claude Sonnet 4 wins:

  • Larger context window
  • Has reasoning mode

GPT-5 wins:

  • Cheaper input tokens
  • Cheaper output tokens
  • Faster response time
  • Higher intelligence benchmark
  • Better at coding
  • Better at math
Price Advantage
GPT-5
Benchmark Advantage
GPT-5
Context Window
Claude Sonnet 4
Speed
GPT-5

Pricing Comparison

Price Comparison

MetricClaude Sonnet 4GPT-5Winner
Input (per 1M tokens)$3.00$1.25 GPT-5
Output (per 1M tokens)$15.00$10.00 GPT-5
Cache Read (per 1M)$300000.00$125000.00 GPT-5
Cache Write (per 1M)$3750000.00N/A Claude Sonnet 4
Using a 3:1 input/output ratio, GPT-5 is 43% cheaper overall.

Claude Sonnet 4 Providers

Amazon Bedrock $3.00 (Cheapest)
OpenRouter $3.00 (Cheapest)
Google $3.00 (Cheapest)
Anthropic $3.00 (Cheapest)

GPT-5 Providers

OpenAI $1.25 (Cheapest)
OpenRouter $1.25 (Cheapest)
Azure $1.25 (Cheapest)

Benchmark Comparison

8
Benchmarks Compared
0
Claude Sonnet 4 Wins
8
GPT-5 Wins

Benchmark Scores

BenchmarkClaude Sonnet 4GPT-5Winner
Intelligence Index
Overall intelligence score
32.744.1
Coding Index
Code generation & understanding
29.434.6
Math Index
Mathematical reasoning
38.094.3
MMLU Pro
Academic knowledge
83.787.1
GPQA
Graduate-level science
68.385.4
LiveCodeBench
Competitive programming
44.984.6
Aider
Real-world code editing
61.388.0
AIME
Competition math
40.795.7
GPT-5 significantly outperforms in coding benchmarks.

Cost vs Quality

X-axis:
Y-axis:
Loading chart...
Claude Sonnet 4
Other models

Context & Performance

Context Window

Claude Sonnet 4
1,000,000
tokens
Max output: 64,000 tokens
GPT-5
400,000
tokens
Max output: 128,000 tokens
Claude Sonnet 4 has a 60% larger context window.

Speed Performance

MetricClaude Sonnet 4GPT-5Winner
Tokens/second66.3 tok/s115.8 tok/s
Time to First Token1.90s100.46s
GPT-5 responds 75% faster on average.

Capabilities

Feature Comparison

FeatureClaude Sonnet 4GPT-5
Vision (Image Input)
Tool/Function Calls
Reasoning Mode
Audio Input
Audio Output
PDF Input
Prompt Caching
Web Search

License & Release

PropertyClaude Sonnet 4GPT-5
LicenseProprietaryProprietary
AuthorAnthropicOpenAI
ReleasedMay 2025Aug 2025

Claude Sonnet 4 Modalities

Input
imagetextfile
Output
text

GPT-5 Modalities

Input
textimagefile
Output
text

Related Comparisons

Compare Claude Sonnet 4 with:

Compare GPT-5 with:

Frequently Asked Questions

GPT-5 has cheaper input pricing at $1.25/M tokens. GPT-5 has cheaper output pricing at $10.00/M tokens.
GPT-5 scores higher on coding benchmarks with a score of 34.6, compared to Claude Sonnet 4's score of 29.4.
Claude Sonnet 4 has a 1,000,000 token context window, while GPT-5 has a 400,000 token context window.
Claude Sonnet 4 supports vision. GPT-5 supports vision.