Head-to-head comparison

Claude Opus 4.5 vs. GPT-5.5

Compare published benchmark results from matching versions and cohorts, API pricing, and technical specifications. Any documented test setup differences remain visible.

shared result series: 24
distinct benchmarks: 14
latest retrieval: 2026-09-29

First modelSecond model

Specifications

General specifications

Release date, availability, context window, architecture, and published parameter counts at a glance. Undisclosed details remain marked as unavailable.

Specifications for Claude Opus 4.5 and GPT-5.5
Attribute	Claude Opus 4.5	GPT-5.5
Provider	Anthropic	OpenAI
Released	Nov 2025	Apr 23, 2026
Availability	Active	Active
Model type	Proprietary	Proprietary
Parameters	Not published	Not published
Architecture	Not published	Not published
Context window	200,000 tokens	1,050,000 tokens
Knowledge cutoff	May 2025	December 1, 2025
Technical sources	Model, Context	Model

Pricing

API pricing compared

The table compares official standard API rates per 1M tokens, including published cache pricing. Batch, fast, priority, regional surcharges, tool calls, and cloud platform pricing are excluded.

API pricing for Claude Opus 4.5 and GPT-5.5
Price type	Claude Opus 4.5	GPT-5.5
API input	$5	$5 ≤272K / $10 >272K
API output	$25	$30 ≤272K / $45 >272K
Cache read	$0.5	$0.5 ≤272K / $1 >272K
Cache write (5 min.)	$6.25	Not listed
Cache write (1 hr.)	$10	Not listed
Price verified	09/22/2026 Source	09/22/2026 Source

Claude Opus 4.5GPT-5.5

API inputUSD per 1M tokens, base tier

Claude Opus 4.5$5

GPT-5.5$5

API outputUSD per 1M tokens, base tier

Claude Opus 4.5$25

GPT-5.5$30

Cache readUSD per 1M tokens, base tier

Claude Opus 4.5$0.5

GPT-5.5$0.5

Key to the marks:better valuebetter, but not conclusivetie

Pricing by provider

Direct developer pricing and provider endpoint tariffs routed through OpenRouter are listed separately. Prices are in USD per 1M tokens. Context-dependent direct tiers plus regional, Flex, Priority, and other OpenRouter tariffs remain individually identifiable.

Provider pricing for Claude Opus 4.5 and GPT-5.5
Model	Provider	Purchase route and tariff	Input per 1M tokens	Output per 1M tokens	Source
Claude Opus 4.5	Anthropic	Direct from developerStandard API	Lowest listed price: $5	Lowest listed price: $25	Source09/22/2026
	Amazon Bedrock	Via OpenRouteramazon-bedrock	Lowest listed price: $5	Lowest listed price: $25	Source10/07/2026
	Amazon Bedrock	Via OpenRouteramazon-bedrock/eu-west-1	$5.5	$27.5	Source10/07/2026
	Anthropic	Via OpenRouteranthropic	Lowest listed price: $5	Lowest listed price: $25	Source10/07/2026
	Azure	Via OpenRouterazure/global	Lowest listed price: $5	Lowest listed price: $25	Source10/07/2026
	Claude Platform on AWS	Via OpenRouterclaude-on-aws	Lowest listed price: $5	Lowest listed price: $25	Source10/07/2026
	Google	Via OpenRoutergoogle-vertex/global	Lowest listed price: $5	Lowest listed price: $25	Source10/07/2026
GPT-5.5	OpenAI	Direct from developerUp to 272,000 context tokens	$5	$30	Source09/22/2026
	OpenAI	Direct from developerAbove 272,000 context tokens	$10	$45	Source09/22/2026
	Amazon Bedrock	Via OpenRouteramazon-bedrock/us-east-1	$5.5	$33	Source10/07/2026
	Azure	Via OpenRouterazure	$5	$30	Source10/07/2026
	Azure	Via OpenRouterazure/eu	$5.5	$33	Source10/07/2026
	Azure	Via OpenRouterazure/us	$5.5	$33	Source10/07/2026
	OpenAI	Via OpenRouteropenai	$5	$30	Source10/07/2026
	OpenAI	Via OpenRouteropenai/fast	$12.5	$75	Source10/07/2026
	OpenAI	Via OpenRouteropenai/flex	Lowest listed price: $2.5	Lowest listed price: $15	Source10/07/2026

Performance

Shared benchmarks

Only published results with the same benchmark version, task, metric, and comparison cohort are paired. Different reasoning levels, agents, harnesses, or output limits appear directly in the table.

Capability profile from matched benchmarks

Each axis averages directly matched benchmark families. A value of 100 means the stronger value within this pair, not a universal quality score.

Claude Opus 4.5

GPT-5.5

Coding: 88 / 100 (1 benchmark family)

Reasoning: 85.4 / 100 (1 benchmark family)

Knowledge: 97.1 / 100 (1 benchmark family)

Finance: 96.2 / 100 (3 benchmark families)

Legal: 95.7 / 100 (1 benchmark family)

Agentic tasks: 93 / 97.2 (3 benchmark families)

Multimodal: 91.9 / 100 (1 benchmark family)

Health: 93.9 / 100 (2 benchmark families)

24 matched rows from 14 benchmark families were considered. The radar shows 8 of 9 comparable categories.

Claude Opus 4.5GPT-5.5

CorpFin v2Overall. Higher is better. No documented setup difference

Claude Opus 4.561.305 %

GPT-5.568.415 %

GAIA ReliabilityAccuracy. Higher is better. No documented setup difference

Claude Opus 4.568.5 %

GPT-5.562.8 %

GPQA DiamondOverall. Higher is better. No documented setup difference

Claude Opus 4.579.546 %

GPT-5.593.182 %

LegalBenchOverall. Higher is better. No documented setup difference

Claude Opus 4.582.837 %

GPT-5.586.515 %

LiveCodeBench v6Overall. Higher is better. No documented setup difference

Claude Opus 4.575.034 %

GPT-5.585.296 %

MedCodeOverall. Higher is better. No documented setup difference

Claude Opus 4.545.174 %

GPT-5.549.1 %

MedScribeOverall. Higher is better. No documented setup difference

Claude Opus 4.583.246 %

GPT-5.586.868 %

MMLU-ProOverall. Higher is better. No documented setup difference

Claude Opus 4.585.59 %

GPT-5.588.144 %

MMMU-ProOverall. Higher is better. No documented setup difference

Claude Opus 4.581.098 %

GPT-5.588.266 %

MortgageTaxOverall. Higher is better. No documented setup difference

Claude Opus 4.568.68 %

GPT-5.568.76 %

SAGEOverall. Higher is better. No documented setup difference

Claude Opus 4.545.002 %

GPT-5.551.532 %

TaxEval v2Overall. Higher is better. No documented setup difference

Claude Opus 4.574.325 %

GPT-5.574.98 %

Key to the marks:better valuebetter, but not conclusivetie

Benchmark scores for Claude Opus 4.5 and GPT-5.5
Benchmark and task	Claude Opus 4.5	GPT-5.5	Source
CorpFin v2Overall Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	61.305 %	Winner68.415 %	Vals AI Source details Data as of: 08/12/2026
GAIA ReliabilityAccuracy Test details No documented difference in the test setupVersion: final leak-cleaned public validation setMetric: exact-match accuracyScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability165 GAIA tasks with repeated runs and perturbations. HAL does not publish cost for this evaluation.	Winner68.5 %	62.8 %	Princeton HAL Source details Data as of: 08/06/2026
GPQA DiamondOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	79.546 %	Winner93.182 %	Vals AI Source details Data as of: 09/01/2026
LegalBenchOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	82.837 %	Winner86.515 %	Vals AI Source details Data as of: 09/27/2026
LiveCodeBench v6Overall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	75.034 %	Winner85.296 %	Vals AI Source details Data as of: 09/01/2026
MedCodeOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	45.174 %	Winner49.1 %	Vals AI Source details Data as of: 09/26/2026
MedScribeOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	83.246 %	Winner86.868 %	Vals AI Source details Data as of: 09/26/2026
MMLU-ProOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	85.59 %	Winner88.144 %	Vals AI Source details Data as of: 09/01/2026
MMMU-ProOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	81.098 %	Winner88.266 %	Vals AI Source details Data as of: 09/01/2026
MortgageTaxOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	68.68 %	Winner68.76 %	Vals AI Source details Data as of: 09/01/2026
SAGEOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	45.002 %	Winner51.532 %	Vals AI Source details Data as of: 09/26/2026
TaxEval v2Overall Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	74.325 %	Winner74.98 %	Vals AI Source details Data as of: 09/01/2026
Terminal-Bench 2.0Overall Test details No documented difference in the test setupVersion: 2.0Metric: accuracyScoring: Higher is betterClaude Opus 4.5: Snapshot: claude-opus-4-5-20251101	58.427 %	Winner73.202 %	Vals AI Source details Data as of: 06/04/2026
τ-bench Airline CleanAccuracy Test details No documented difference in the test setupVersion: 26-task cleaned subsetMetric: task success rateScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability26 cleaned airline tasks. Princeton HAL recomputed every metric from scratch on this subset.	Winner80.8 %	79.5 %	Princeton HAL Source details Data as of: 08/05/2026
GAIA ReliabilityConsistency Test details No documented difference in the test setupVersion: final leak-cleaned public validation setMetric: aggregate consistencyScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability165 GAIA tasks with repeated runs and perturbations. HAL does not publish cost for this evaluation.	Winner0.81 rating points	0.61 rating points	Princeton HAL Source details Data as of: 08/06/2026
τ-bench Airline CleanConsistency Test details No documented difference in the test setupVersion: 26-task cleaned subsetMetric: aggregate consistencyScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability26 cleaned airline tasks. Princeton HAL recomputed every metric from scratch on this subset.	Tie0.84 rating points	Tie0.84 rating points	Princeton HAL Source details Data as of: 08/05/2026
GAIA ReliabilityPredictability Test details No documented difference in the test setupVersion: final leak-cleaned public validation setMetric: aggregate predictabilityScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability165 GAIA tasks with repeated runs and perturbations. HAL does not publish cost for this evaluation.	Winner0.84 rating points	0.8 rating points	Princeton HAL Source details Data as of: 08/06/2026
τ-bench Airline CleanPredictability Test details No documented difference in the test setupVersion: 26-task cleaned subsetMetric: aggregate predictabilityScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability26 cleaned airline tasks. Princeton HAL recomputed every metric from scratch on this subset.	Winner0.85 rating points	0.84 rating points	Princeton HAL Source details Data as of: 08/05/2026
GAIA ReliabilityReliability Test details No documented difference in the test setupVersion: final leak-cleaned public validation setMetric: aggregate reliabilityScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability165 GAIA tasks with repeated runs and perturbations. HAL does not publish cost for this evaluation.	Winner0.85 rating points	0.79 rating points	Princeton HAL Source details Data as of: 08/06/2026
τ-bench Airline CleanReliability Test details No documented difference in the test setupVersion: 26-task cleaned subsetMetric: aggregate reliabilityScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability26 cleaned airline tasks. Princeton HAL recomputed every metric from scratch on this subset.	0.88 rating points	Winner0.89 rating points	Princeton HAL Source details Data as of: 08/05/2026
GAIA ReliabilityRobustness Test details No documented difference in the test setupVersion: final leak-cleaned public validation setMetric: aggregate robustnessScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability165 GAIA tasks with repeated runs and perturbations. HAL does not publish cost for this evaluation.	0.91 rating points	Winner0.95 rating points	Princeton HAL Source details Data as of: 08/06/2026
τ-bench Airline CleanRobustness Test details No documented difference in the test setupVersion: 26-task cleaned subsetMetric: aggregate robustnessScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability26 cleaned airline tasks. Princeton HAL recomputed every metric from scratch on this subset.	Tie0.97 rating points	Tie0.97 rating points	Princeton HAL Source details Data as of: 08/05/2026
GAIA ReliabilitySafety Test details No documented difference in the test setupVersion: final leak-cleaned public validation setMetric: aggregate safetyScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability165 GAIA tasks with repeated runs and perturbations. HAL does not publish cost for this evaluation.	Tie1 rating points	Tie1 rating points	Princeton HAL Source details Data as of: 08/06/2026
τ-bench Airline CleanSafety Test details No documented difference in the test setupVersion: 26-task cleaned subsetMetric: aggregate safetyScoring: Higher is betterSettings: Agent: HAL Reliability, Harness: HAL Reliability26 cleaned airline tasks. Princeton HAL recomputed every metric from scratch on this subset.	Winner0.98 rating points	0.96 rating points	Princeton HAL Source details Data as of: 08/05/2026
Key to the marks:better valuebetter, but not conclusivetie

Our model tests

Our tests

Compare how the models respond to the same prompt. Each test shows the latest attempt, with no subsequent fixes to the generated code. Earlier runs remain available in the test details. These results do not contribute to an overall score.

Pelican on a bicycle

A community classic for free-form SVG drawing.

Claude Opus 4.5

Reasoning (requested): high

Current attempt

Run date: 1 Oct 2026
API provider: openrouter
Requested API model: anthropic/claude-opus-4.5
API model reported in response: anthropic/claude-opus-4.5
Reasoning setting: high
Pinned endpoint: Anthropic | anthropic/claude-4.5-opus-20251124
Temperature: Provider default
Token limit including reasoning: 8,192 tokens
Test protocol: community-visual-tests-v1
Technical checks: Passed, not a quality rating
Response time: 21.6 s

Visually reviewed. This individual test does not establish overall model quality.

GPT-5.5

Reasoning (requested): high

Current attempt

Run date: 1 Oct 2026
API provider: openrouter
Requested API model: openai/gpt-5.5
API model reported in response: openai/gpt-5.5
Reasoning setting: high
Pinned endpoint: OpenAI | openai/gpt-5.5-20260423
Temperature: Provider default
Token limit including reasoning: 8,192 tokens
Test protocol: community-visual-tests-v1
Technical checks: Passed, not a quality rating
Response time: 25.9 s

Visually reviewed. This individual test does not establish overall model quality.

Earlier attempts (1)

7 Sept 2026API request failedReasoning setting: high

Prompt and test conditions

Generate an SVG of a pelican riding a bicycle

Token limit including reasoning: 8,192 tokens

Task origin (Simon Willison)

How to read this comparison

Winning one benchmark is not an overall verdict

Your workload, budget, and required context length matter most. A coding benchmark says little about visual understanding or agent performance.

The standard price table uses direct developer rates. The provider comparison labels OpenRouter endpoints, regions, and special tariffs separately.

Every result links to its measurement source and retrieval date. Labels distinguish vendor reports, official benchmark leaderboards, and independent evaluations. Disputed or archived results are excluded.

More matchups

Compare Claude Opus 4.5 and GPT-5.5 with other leading models using the same data and benchmark logic.

View the complete LLM comparison