The Known Good Updated 29 Jul 2026
Anthropic: Claude Sonnet 4.6 vs OpenAI: GPT-5 ·Anthropic vs OpenAI

Anthropic: Claude Sonnet 4.6 vs OpenAI: GPT-5 — head-to-head comparison

Anthropic: Claude Sonnet 4.6 profile OpenAI: GPT-5 profile All comparisons
Anthropic: Claude Sonnet 4.6 — Anthropic
Quality
#17 of 62
80based on 4 of 7 evaluations
Known Good Index
Speed
#182 of 317
43.9
Output tokens per second
Price
#329 of 370
$6.00
Blended per 1M tokens
In $3.00 · Out $15.00
Cache hit price
#195 of 238
$0.30
per 1M tokens (−90%)
Context
#73 of 375
1.00M
tokens (~1,500 pages)
OpenAI: GPT-5 — OpenAI
Quality
#18 of 62
78based on 6 of 7 evaluations
Known Good Index
Speed
#159 of 317
48.3
Output tokens per second
Price
#289 of 370
$3.44
Blended per 1M tokens
In $1.25 · Out $10.00
Cache hit price
#130 of 238
$0.12
per 1M tokens (−90%)
Context
#136 of 375
400k
tokens (~600 pages)
Summary

On the Known Good Index, Anthropic: Claude Sonnet 4.6 scores 80 and OpenAI: GPT-5 scores 78 — a 2-point lead for Anthropic: Claude Sonnet 4.6. They share 4 published evaluations, with Anthropic: Claude Sonnet 4.6 ahead on 3 and OpenAI: GPT-5 ahead on 1.

Blended pricing is $6.00 per 1M tokens for Anthropic: Claude Sonnet 4.6 and $3.44 for OpenAI: GPT-5 — OpenAI: GPT-5 is about 1.7× cheaper.

Throughput is 44 tokens per second for Anthropic: Claude Sonnet 4.6 and 48 for OpenAI: GPT-5, with time to first token at 1.7s and 2.1s. Anthropic: Claude Sonnet 4.6 accepts 1.00M tokens of context against 400k.

Anthropic: Claude Sonnet 4.6 is proprietary; OpenAI: GPT-5 is proprietary. 5 tracked hosts serve one or both, and 1 serve both.

Generated from tracked data on 29 Jul 2026 · how these comparisons are computed →

At a glance
MetricAnthropic: Claude Sonnet 4.6OpenAI: GPT-5
Known Good Index 80 based on 4 of 7 evaluations 78 based on 6 of 7 evaluations
Blended $/1M $6.00 $3.44
Output tok/s 44 48
Context window 1.00M 400k
Providers 4 2
Jump to section

Quality

Known Good Index and the capability indices, side by side

Index scores

Each index is min-max normalised across tracked models and scaled 0–100 from published evaluations. We do not run these evaluations.

MetricAnthropic: Claude Sonnet 4.6OpenAI: GPT-5
Known Good Index 80 78
Coding Index 71 67
Mathematics Index 86 96
Reasoning Index — 71
Agentic Index 71 67
Openness Index 5 5
The Known Good

Quality Breakdown

Every evaluation both models report a published score for

Evaluations both models report

Bar = Anthropic: Claude Sonnet 4.6 · marker = OpenAI: GPT-5 · scores ingested from Epoch AI (CC-BY 4.0)

GPQA Diamond87 / 86
Mock AIME 2024–2586 / 91
SWE-bench Verified75 / 74
Terminal-Bench53 / 50

Right-hand figures read Anthropic: Claude Sonnet 4.6 / OpenAI: GPT-5. Anthropic: Claude Sonnet 4.6 leads on 3 of 4, OpenAI: GPT-5 on 1.

The Known Good

Arena Elo

Human preference rating from blind pairwise votes

Arena ratings

Elo from blind pairwise votes, ingested from Arena. Higher is better.

MetricAnthropic: Claude Sonnet 4.6OpenAI: GPT-5
Text Arena 1,457 —
Text Arena (style controlled) 1,472 —
The Known Good

Price & Cost

Published API pricing by token type

Published pricing

USD per 1M tokens. Blended is a 3:1 input:output mix. Live from OpenRouter, cross-checked against models.dev. Lower is better.

MetricAnthropic: Claude Sonnet 4.6OpenAI: GPT-5
Input $/1M $3.00 $1.25
Output $/1M $15.00 $10.00
Blended $/1M $6.00 $3.44
Cache read $/1M $0.300 $0.125
The Known Good
We publish blended price rather than cost-per-task. Cost-per-task requires token counts from an evaluation suite we do not run. Blended price is computed from published rates and is fully reproducible — see the pricing methodology.

Context Window

Maximum input accepted and maximum output emitted

Context and output limits

Provider-declared limits, from metadata via models.dev and OpenRouter. Higher is better.

MetricAnthropic: Claude Sonnet 4.6OpenAI: GPT-5
Context window 1.00M 400k
Max output tokens 128k 128k
The Known Good

Speed & Latency

Median throughput and time to first token across hosting providers

Output speedTime to first token
Throughput and latency

Medians across all tracked providers, derived from OpenRouter endpoint telemetry.

MetricAnthropic: Claude Sonnet 4.6OpenAI: GPT-5
Output tokens/s 44 48
Time to first token 1.7s 2.1s
The Known Good

Providers

Which hosts serve each model, and at what price and speed

5 hosts · 1 serve both
Full provider benchmarks →
Provider Anthropic: Claude Sonnet 4.6 blendedAnthropic: Claude Sonnet 4.6 tok/s OpenAI: GPT-5 blendedOpenAI: GPT-5 tok/s
Google · first party $6.00 52 — —
Anthropic · first party $6.00 38 — —
Azure · first party $6.00 38 $3.44 46
Amazon Bedrock · first party $6.00 36 — —
OpenAI · first party — — $3.44 67

Per-host pricing and telemetry from OpenRouter. A dash means that host does not serve that model.

The Known Good
Sources for this page
Pricing & providers — OpenRouter Metadata — models.dev Benchmarks — Epoch AI (CC-BY 4.0)Elo — Arena Full attribution →