Compare · ModelsLive · 2 picked · head to head
GPT-5.3-Codex vs Claude Sonnet 4.6
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GPT-5.3-Codex wins on 5/7 benchmarks
GPT-5.3-Codex wins 5 of 7 shared benchmarks. Leads in speed · knowledge · agentic.
Category leads
speed·GPT-5.3-Codexknowledge·GPT-5.3-Codexagentic·GPT-5.3-Codexcoding·Claude Sonnet 4.6
Hype vs Reality
Attention vs performance
GPT-5.3-Codex
#84 by perf·no signal
Claude Sonnet 4.6
#102 by perf·#18 by attention
Best value
GPT-5.3-Codex
1.3x better value than Claude Sonnet 4.6
GPT-5.3-Codex
6.6 pts/$
$7.88/M
Claude Sonnet 4.6
5.3 pts/$
$9.00/M
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
Anthropic
$380.0B·Tier 1
Head to head
7 benchmarks · 2 models
GPT-5.3-CodexClaude Sonnet 4.6
Artificial Analysis · Agentic Index
Claude Sonnet 4.6 leads by +0.8
GPT-5.3-Codex
62.2
Claude Sonnet 4.6
63.0
Artificial Analysis · Coding Index
GPT-5.3-Codex leads by +2.2
GPT-5.3-Codex
53.1
Claude Sonnet 4.6
50.9
Artificial Analysis · Quality Index
GPT-5.3-Codex leads by +2.3
GPT-5.3-Codex
54.0
Claude Sonnet 4.6
51.7
PostTrainBench
GPT-5.3-Codex leads by +1.3
GPT-5.3-Codex
17.8
Claude Sonnet 4.6
16.4
SWE Atlas · Codebase QnA
GPT-5.3-Codex leads by +1.4
GPT-5.3-Codex
32.6
Claude Sonnet 4.6
31.2
SWE-Bench verified
Claude Sonnet 4.6 leads by +0.4
GPT-5.3-Codex
74.8
Claude Sonnet 4.6
75.2
WeirdML
GPT-5.3-Codex leads by +13.2
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GPT-5.3-Codex
79.3
Claude Sonnet 4.6
66.1
Full benchmark table
| Benchmark | GPT-5.3-Codex | Claude Sonnet 4.6 |
|---|---|---|
Artificial Analysis · Agentic Index | 62.2 | 63.0 |
Artificial Analysis · Coding Index | 53.1 | 50.9 |
Artificial Analysis · Quality Index | 54.0 | 51.7 |
PostTrainBench | 17.8 | 16.4 |
SWE Atlas · Codebase QnA | 32.6 | 31.2 |
SWE-Bench verified | 74.8 | 75.2 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 79.3 | 66.1 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| GPT-5.3-Codex | $1.75 | $14.00 | 400K tokens (~200 books) | $48.13 |
| Claude Sonnet 4.6 | $3.00 | $15.00 | 1.0M tokens (~500 books) | $60.00 |