CompareLLM.ai
Live
Leaderboard
Models
Compare
Best of & Stacks
Research & News
…
CompareLLM.ai
Precision Benchmarks

Programmatic, dated AI model benchmarks, head-to-head comparisons, and Stack Engine presets.

Daily ingest · 06:00 UTC

Analytics & Benchmarks

  • AI Model Leaderboard
  • Head-to-Head Compare Hub
  • Models Directory
  • Stack Engine Presets
  • Frontier Models
  • Open Weights Catalog

Guides & Intent Lists

  • Best LLM Lists (2026)
  • Best Coding LLM
  • Best Cheap LLM
  • Fastest Low-Latency LLM
  • Claude vs GPT Benchmark
  • What is Elo?
  • Methodology Guides
  • News & Dispatches

Transparency & API

  • Evaluation Methodology
  • Benchmark Changelog
  • Public JSON API
  • llms.txt Specification
  • Privacy Policy
  • Sign In / Account

© 2026 CompareLLM. Public benchmark data aggregated from Arena Elo, LiveBench, SWE-bench & OpenRouter.

Every score has a dated snapshot.

Theme:
Currency:
  1. Home
  2. Comparisons
  3. Claude Sonnet 4 vs GPT-5

Pairwise benchmark snapshot · Aug 1, 2026

Claude Sonnet 4 vs GPT-5 benchmark

In this head-to-head showdown, GPT-5 delivers higher overall intelligence and human-preferred responses (Elo 1,558 vs 1,455), while GPT-5 leads in SWE-bench software engineering benchmarks (68.4%), while Claude Sonnet 4 is more budget-friendly at $15/1M per 1M output tokens. Review the complete breakdown below to determine which model best fits your performance and budget requirements.

Executive Comparison Verdict

Bottom Line: Choose GPT-5 if your priority is peak reasoning, complex code generation, and top preference Elo. Choose Claude Sonnet 4 if you are optimizing for low latency, high throughput, and cost-efficient API deployment.
Share Analysis:
WhatsAppTelegramXLinkedInReddit

Verified Head-to-Head Advantage Breakdown

10 direct benchmark disciplines evaluated across capability, speed, and cost

Claude Sonnet 4 (4)GPT-5 (6)
Claude Sonnet 4
4 of 10 Wins
✓Time to first token✓Output speed✓Input price✓Output price
GPT-5
6 of 10 Wins
✓Preference Elo✓Coding Elo✓LiveBench✓SWE-bench✓GPQA Diamond✓Context window
Anthropic
Leads 4 of 10 metrics
Claude Sonnet 4

First Sonnet 4 generation. Bridge between 3.5/3.7 and Sonnet 5.

Elo 1,455$15/1M out
OpenAI
Leads 6 of 10 metrics
GPT-5

OpenAI flagship reasoning model for 2025–26. Strong general preference Elo and multimodal coverage.

Elo 1,558$20/1M out

GPT-5 Thinking Depth & Cost Scaler

Model B

Simulate accuracy gains vs added response time & token cost

Mode: Medium (Standard)

Reasoning models “think before answering” by generating internal reasoning tokens. Higher effort improves math, coding, and logic accuracy, but increases response delay and token costs.

Thinking Delay
290 ms
Standard speed
Reasoning Depth
~1,500
Internal tokens
Cost / 1k Calls
$38.50
4.5× standard bill
Accuracy Boost
+4.5% accuracy
STEM & SWE-bench
Top Rival Showdowns for Claude Sonnet 4
All Matchups
Compare vs:vs GPT-5 minivs Gemini 3 Provs Grok 4.6vs DeepSeek V4 Provs Qwen 3 Max
Preference LeaderGPT-5 Wins

GPT-5

Δ 103 Arena Elo pts

Throughput LeaderClaude Sonnet 4 Wins

Claude Sonnet 4

92 tok/s

Value per Dollar LeaderClaude Sonnet 4 Wins

Claude Sonnet 4

$15 / 1M output

Capability Percentiles

Relative percentile scores computed across all active models in the benchmark catalog.

Multi-Dimensional Capability Radar

Leaders Matchup Capability Radar

Comparing top Western standard models with China's leading frontier rival across 6 skill dimensions. Tap any spoke or dot to inspect.

Tap any node to inspect
Percentile 0–100
Head-to-Head Comparison

Overall Matchup Breakdown

Category wins across reasoning intelligence, generation speed, and token cost.

Claude Sonnet 4 (4)vsGPT-5 (6)
Claude Sonnet 4: 4W (40%)Overall: GPT-5GPT-5: 6W (60%)
← Claude Sonnet 4GPT-5 →
🏆GPT-5(5/5)

Intelligence & Reasoning

Preference Elo, Coding proficiency, SWE-bench & LiveBench accuracy

Benchmark
Claude Sonnet 4vsGPT-5
Preference Elo
1,455vs1,558+103
Coding Elo
1,468vs1,540+72
SWE-bench
62.8%vs68.4%+5.6%
LiveBench
63.2%vs69.8%+6.6%
GPQA Diamond
76.4%vs85.2%+8.8%
Claude Sonnet 4: 0WGPT-5: 5W
🏆Claude Sonnet 4(2/2)

Speed & Latency

Generation throughput and time to first token responsiveness

Benchmark
Claude Sonnet 4vsGPT-5
Output speed
92 tok/svs78 tok/s+14 tok/s
Time to first token
220 msvs290 ms+70 ms
Claude Sonnet 4: 2WGPT-5: 0W
🏆Claude Sonnet 4(2/3)

Pricing & Capacity

Cost per million tokens and max context window length

Benchmark
Claude Sonnet 4vsGPT-5
Output price
$15/1Mvs$20/1M+$5/1M
Input price
$3/1Mvs$5/1M+$2/1M
Context window
200kvs400k+200k
Claude Sonnet 4: 2WGPT-5: 1W

Side-by-Side Benchmark Matrix

Preference Elo+103
Claude Sonnet 41,455
seed-bootstrap · Aug 1, 2026
GPT-51,558
seed-bootstrap · Aug 1, 2026
Coding Elo+72
Claude Sonnet 41,468
seed-bootstrap · Aug 1, 2026
GPT-51,540
seed-bootstrap · Aug 1, 2026
LiveBench+6.6%
Claude Sonnet 463.2%
seed-bootstrap · Aug 1, 2026
GPT-569.8%
seed-bootstrap · Aug 1, 2026
SWE-bench+5.6%
Claude Sonnet 462.8%
seed-bootstrap · Aug 1, 2026
GPT-568.4%
seed-bootstrap · Aug 1, 2026
GPQA Diamond+8.8%
Claude Sonnet 476.4%
seed-bootstrap · Aug 1, 2026
GPT-585.2%
seed-bootstrap · Aug 1, 2026
Time to first token+70 ms
Claude Sonnet 4220 ms
seed-bootstrap · Aug 1, 2026
GPT-5290 ms
seed-bootstrap · Aug 1, 2026
Output speed+14 tok/s
Claude Sonnet 492 tok/s
seed-bootstrap · Aug 1, 2026
GPT-578 tok/s
seed-bootstrap · Aug 1, 2026
Input price+$2/1M
Claude Sonnet 4$3/1M
seed-bootstrap · Aug 1, 2026
GPT-5$5/1M
seed-bootstrap · Aug 1, 2026
Output price+$5/1M
Claude Sonnet 4$15/1M
seed-bootstrap · Aug 1, 2026
GPT-5$20/1M
seed-bootstrap · Aug 1, 2026
Context window+200k
Claude Sonnet 4200k
seed-bootstrap · Aug 1, 2026
GPT-5400k
seed-bootstrap · Aug 1, 2026
BenchmarkClaude Sonnet 4GPT-5Advantage Delta
Preference Elo
1,455
seed-bootstrap · Aug 1, 2026
1,558
seed-bootstrap · Aug 1, 2026
+103
Coding Elo
1,468
seed-bootstrap · Aug 1, 2026
1,540
seed-bootstrap · Aug 1, 2026
+72
LiveBench
63.2%
seed-bootstrap · Aug 1, 2026
69.8%
seed-bootstrap · Aug 1, 2026
+6.6%
SWE-bench
62.8%
seed-bootstrap · Aug 1, 2026
68.4%
seed-bootstrap · Aug 1, 2026
+5.6%
GPQA Diamond
76.4%
seed-bootstrap · Aug 1, 2026
85.2%
seed-bootstrap · Aug 1, 2026
+8.8%
Time to first token
220 ms
seed-bootstrap · Aug 1, 2026
290 ms
seed-bootstrap · Aug 1, 2026
+70 ms
Output speed
92 tok/s
seed-bootstrap · Aug 1, 2026
78 tok/s
seed-bootstrap · Aug 1, 2026
+14 tok/s
Input price
$3/1M
seed-bootstrap · Aug 1, 2026
$5/1M
seed-bootstrap · Aug 1, 2026
+$2/1M
Output price
$15/1M
seed-bootstrap · Aug 1, 2026
$20/1M
seed-bootstrap · Aug 1, 2026
+$5/1M
Context window
200k
seed-bootstrap · Aug 1, 2026
400k
seed-bootstrap · Aug 1, 2026
+200k
Interactive Simulator (USD)

Workload Cost & Savings Calculator

Simulate monthly production API costs in USD (US Dollar).

Save up to 31% with Claude Sonnet 4
Monthly Volume50M tokens/mo
Quick Presets:
Token Ratio (In vs Out)70% In / 30% Out
RAG / Search (10% out)Coding / Chat (50% out)
Claude Sonnet 4$330.00 / mo
In: $105Out: $225
GPT-5$475.00 / mo
In: $175Out: $300
Estimated Cost Delta

Claude Sonnet 4 is estimated to save $145.00/month ($1,740/year).

Based on 50M tokens (30% gen)

Key Selection Recommendations

  • 1
    GPT-5:Pick GPT-5 when repo-level coding accuracy is the constraint.
  • 2
    Claude Sonnet 4:Pick Claude Sonnet 4 when you are optimizing output cost.
  • 3
    Claude Sonnet 4:Pick Claude Sonnet 4 when time-to-first-token matters more than peak Elo.

Recommended Workload Routing

Repo / coding agentsGPT-5

Higher SWE-bench (68.4%).

High-volume chatClaude Sonnet 4

Lower output list price ($15/1M).

Voice / low-latency UIClaude Sonnet 4

Lower TTFT (220 ms).

Long-document RAGGPT-5

Larger window (400k).

Screenshots / visionGPT-5

Both accept images. Defaulting to the higher-Elo side (GPT-5).

Target WorkloadRecommended PickEvaluation Rationale
Repo / coding agentsGPT-5Higher SWE-bench (68.4%).
High-volume chatClaude Sonnet 4Lower output list price ($15/1M).
Voice / low-latency UIClaude Sonnet 4Lower TTFT (220 ms).
Long-document RAGGPT-5Larger window (400k).
Screenshots / visionGPT-5Both accept images. Defaulting to the higher-Elo side (GPT-5).

Related Editorial Dispatches & Benchmark Notes

View all news
launchAug 8, 2025

GPT-5 (2025) remains the legacy OpenAI flagship baseline

Still a common compare target. 5.6 Sol replaced it as the buy; GPT-5 stays for old URLs.

Community Sentiment

Cast Your Matchup Vote

0 total votes

Community Discussions (0)

Sign in to cast your verified vote, bookmark models, and participate in benchmark discussions.Sign In to Post

No comments posted on this matchup yet. Be the first to share an evaluation note!

Social Share Card Preview

Dynamic OpenGraph banner generated at /compare/claude-sonnet-4-vs-gpt-5/opengraph-image

Save PNG

CompareLLM AI Matrix

Verified Head-to-Head Telemetry

LIVE BENCHMARK
VS
Anthropic

Claude Sonnet 4

Elo 1,455
SWE-bench62.8%
Speed (tok/s)92 tok/s
Input / 1M Tokens$3/1M
Output / 1M Tokens$15/1M
OpenAI

GPT-5

Elo 1,558
SWE-bench68.4%
Speed (tok/s)78 tok/s
Input / 1M Tokens$5/1M
Output / 1M Tokens$20/1M

Frequently Asked Questions

Which is better overall, Claude Sonnet 4 or GPT-5?
GPT-5 has the higher preference Elo in our latest snapshot (1,558). “Better” still depends on coding, price, and latency — see the table.
Which is better at coding, Claude Sonnet 4 or GPT-5?
GPT-5 leads SWE-bench at 68.4% vs 62.8%. SWE-bench is one harness, not your repo.
Which is cheaper to run in production?
Claude Sonnet 4 output tokens are $15/1M versus $20/1M. Input prices and retry rates still move the real bill.
Which is faster, Claude Sonnet 4 or GPT-5?
Claude Sonnet 4 has the lower time-to-first-token (220 ms vs 290 ms). Tokens/sec is a separate column if you care about long completions.
Which has the larger context window?
GPT-5 accepts 400k tokens versus 200k.
Can I self-host either model?
Neither model is marked open-weights here. You are comparing hosted APIs.
Are these scores from CompareLLM’s own evals?
No. V1 aggregates public leaderboards (and optional first-party latency pings). Each cell names its source and date. Read /methodology.
What does preference Elo mean on this page?
Preference Elo is a crowd vote from LMArena / Arena. People see two hidden answers and pick the one they like more. The model that wins more often gets a higher Elo. That means people preferred it — not that it passed a school test. It is not SWE-bench, not accuracy, and not a number we invent. Claude Sonnet 4 and GPT-5 Elo cells are dated snapshots from the named source — we do not compute Arena ourselves.

Related news

  • launch · Aug 8, 2025

    GPT-5 (2025) remains the legacy OpenAI flagship baseline

Similar Head-to-Head Comparisons

All Claude Sonnet 4 matchups →|All GPT-5 matchups →
Claude Sonnet 4 vs Claude 3.7 Sonnet (previous sonnet)Claude Sonnet 4 vs Claude Sonnet 4.5 (next sonnet)GPT-5 vs GPT-4.5 Orion (previous gpt-flagship)GPT-5 vs GPT-5.6 Sol (next gpt-flagship)Claude Sonnet 4.5 vs GPT-5GPT-5.6 Terra vs GPT-5