Claude Sonnet 4.6 vs GPT-5
Complete pricing and performance comparison between Anthropic's Claude Sonnet 4.6 and OpenAI's GPT-5.
Quick Verdict
Cheaper
GPT-5
2.4x cheaper input, 1.5x cheaper output
Larger Context
GPT-5
400K vs 200K
Higher Quality
GPT-5
Score: 45 vs 44
Faster
GPT-5
60 vs 54 tok/s
Pricing Comparison
| Spec | Claude Sonnet 4.6 | GPT-5 | Difference |
|---|---|---|---|
| Provider | Anthropic | OpenAI | |
| Input / 1M tokens | $3 | $1.25 | GPT-5 is 58% more expensive |
| Output / 1M tokens | $15 | $10 | GPT-5 is 33% more expensive |
| Context Window | 200K | 400K | 2x difference |
| Max Output | 64K | 16K | |
| Tokenizer | cl100k_base | o200k_base |
Performance Benchmarks
| Metric | Claude Sonnet 4.6 | GPT-5 | Winner |
|---|---|---|---|
| Quality Index | 44 | 45 | GPT-5 |
| Output Speed | 54 tok/s | 60 tok/s | GPT-5 |
| Time to First Token | 1.17s | 117.69s | Claude Sonnet 4.6 |
| Value (Quality/$) | 14.8 | 35.7 | Higher = better value |
Benchmark data from Artificial Analysis. Quality Index is a composite score across reasoning, coding, and knowledge tasks.
Cost at Scale
Estimated cost at different usage levels (3:1 input-to-output token ratio, typical for chat).
| Usage | Claude Sonnet 4.6 | GPT-5 | Savings |
|---|---|---|---|
Single request 1K in / 300 out | $0.0075 | $0.0042 | GPT-5 saves $0.0033 |
10 requests 10K in / 3K out | $0.075 | $0.042 | GPT-5 saves $0.033 |
100 requests 100K in / 30K out | $0.750 | $0.425 | GPT-5 saves $0.325 |
1,000 requests 1M in / 300K out | $7.50 | $4.25 | GPT-5 saves $3.25 |
10,000 requests 10M in / 3M out | $75.00 | $42.50 | GPT-5 saves $32.50 |
1M requests/mo 1B in / 300M out | $7500.00 | $4250.00 | GPT-5 saves $3250.00 |
Pros & Cons
Claude Sonnet 4.6 Strengths
- +Higher max output tokens
- +Lower latency (faster first token)
GPT-5 Strengths
- +Cheaper input tokens
- +Cheaper output tokens
- +Larger context window (400K vs 200K)
- +Faster output (60 vs 54 tok/s)
- +Higher quality score (45 vs 44)
When to Use Each Model
Choose Claude Sonnet 4.6 for
- →Generating long-form content or detailed code
Choose GPT-5 for
- →Budget-conscious projects where cost is the primary factor
- →Long documents, large codebases, or multi-turn conversations
- →Tasks requiring maximum accuracy and reasoning
- →Real-time applications, chat, or autocomplete
Frequently Asked Questions
Which is cheaper, Claude Sonnet 4.6 or GPT-5?
For input tokens, GPT-5 is 2.4x cheaper at $1.25/1M tokens. For output tokens, GPT-5 is 1.5x cheaper at $10/1M tokens. At typical usage (1M input + 300K output), Claude Sonnet 4.6 costs $7.50 vs GPT-5 at $4.25.
What's the context window difference?
Claude Sonnet 4.6 supports 200K context (200,000 tokens), while GPT-5 supports 400K (400,000 tokens). GPT-5 can handle 2x more context in a single request.
Which model has better benchmarks?
Quality Index: Claude Sonnet 4.6 scores 44 vs GPT-5 at 45. Speed: Claude Sonnet 4.6 generates 54 tok/s vs GPT-5 at 60 tok/s. Time to first token: Claude Sonnet 4.6 at 1.17s vs GPT-5 at 117.69s.
When should I choose Claude Sonnet 4.6 over GPT-5?
Choose Claude Sonnet 4.6 when you need: Higher max output tokens, Lower latency (faster first token). Choose GPT-5 when you need: Cheaper input tokens, Cheaper output tokens, Larger context window (400K vs 200K), Faster output (60 vs 54 tok/s), Higher quality score (45 vs 44).
How much would 10,000 API requests cost?
At 1K input + 300 output tokens per request (typical chat): Claude Sonnet 4.6 = $75.00, GPT-5 = $42.50. At 10K input + 1K output per request (longer conversations): Claude Sonnet 4.6 = $450.00, GPT-5 = $225.00.