Skip to main content
BenchLM

Model comparison

Claude Opus 4.8 vs GPT-5.3 Codex

Data verified

Head-to-head evidence from 18 shared benchmark results across 6 categories. Overall scores shown here use BenchLM's provisional ranking lane.

85/100
Margin
3.0pts
← winning
82/100
2 category wins0 category wins

Verified leaderboard positions: Claude Opus 4.8 #3; GPT-5.3 Codex unranked

Evidence parity. Claude Opus 4.8 and GPT-5.3 Codex share 18 comparable benchmark results. 2 of 8 categories are comparable. 35 results are unique to Claude Opus 4.8; 4 to GPT-5.3 Codex.

Updated July 12, 2026
Shared results
18
Claude Opus 4.8 only
35
GPT-5.3 Codex only
4
Comparable categories
2 / 8

Pick Claude Opus 4.8 if you want the stronger benchmark profile. GPT-5.3 Codex only becomes the better choice if you want the cheaper token bill.

Confidence note. This is a partial-evidence comparison with 18 shared benchmark results across 6 evidence categories; 2 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Claude Opus 4.8 has the cleaner provisional overall profile here, landing at 85 versus 82. It is a real lead, but still close enough that category-level strengths matter more than the headline number.

Claude Opus 4.8's sharpest advantage is in coding, where it averages 76.4 against 64.4. The single biggest benchmark swing on the page is OSWorld-Verified, 83.4% to 64.7%.

Claude Opus 4.8 is also the more expensive model on tokens at $5.00 input / $25.00 output per 1M tokens, versus $1.75 input / $14.00 output per 1M tokens for GPT-5.3 Codex. Claude Opus 4.8 gives you the larger context window at 1M, compared with 400K for GPT-5.3 Codex.

Category breakdown

Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.

Category scores and score margins for Claude Opus 4.8 and GPT-5.3 Codex
CategoryClaude Opus 4.8ΔGPT-5.3 Codex
CodingClaude Opus 4.876.4Margin← 12.0GPT-5.3 Codex64.4
AgenticClaude Opus 4.880.3Margin← 8.9GPT-5.3 Codex71.4
KnowledgeClaude Opus 4.862.7MarginNo overlapGPT-5.3 CodexNot measured
MathClaude Opus 4.853.9MarginNo overlapGPT-5.3 CodexNot measured
MultimodalClaude Opus 4.877.0MarginNo overlapGPT-5.3 CodexNot measured

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · Claude Opus 4.8B · GPT-5.3 Codex

  1. OSWorld-Verified

    Agentic
    Source ↗
    A 83.4%B 64.7%
    Winner: Claude Opus 4.8Δ 18.7
    OSWorld-Verified: Claude Opus 4.8 scored 83.4%; GPT-5.3 Codex scored 64.7%. Claude Opus 4.8 wins this benchmark.

  2. SWE-bench Pro

    Coding
    Source ↗
    A 69.2%B 56.8%
    Winner: Claude Opus 4.8Δ 12.4
    SWE-bench Pro: Claude Opus 4.8 scored 69.2%; GPT-5.3 Codex scored 56.8%. Claude Opus 4.8 wins this benchmark.

  3. SWE-bench Verified

    Coding
    Source ↗
    A 88.6%B 85%
    Winner: Claude Opus 4.8Δ 3.6
    SWE-bench Verified: Claude Opus 4.8 scored 88.6%; GPT-5.3 Codex scored 85%. Claude Opus 4.8 wins this benchmark.

  4. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 74.6%B 77.3%
    Winner: GPT-5.3 CodexΔ 2.7
    Terminal-Bench 2.0: Claude Opus 4.8 scored 74.6%; GPT-5.3 Codex scored 77.3%. GPT-5.3 Codex wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricClaude Opus 4.8GPT-5.3 CodexComparison
Input / output priceUSD per 1M tokensClaude Opus 4.8$5 input / $25 outputGPT-5.3 Codex$1.75 input / $14 outputGPT-5.3 Codex has the lower combined listed price.
Generation speedtokens per secondClaude Opus 4.8Not availableGPT-5.3 Codex79 tok/sA complete speed comparison is not available.
First-answer latencyseconds to first tokenClaude Opus 4.8Not availableGPT-5.3 Codex88.26 sA complete latency comparison is not available.
Context windowmaximum listed tokensClaude Opus 4.81MGPT-5.3 Codex400KClaude Opus 4.8 lists the larger context window.

Benchmark Deep Dive

AgenticClaude Opus 4.8 wins
BenchmarkClaude Opus 4.8GPT-5.3 CodexResult
Terminal-Bench 2.0Source 74.6%77.3%GPT-5.3 Codex leads
BrowseCompSource 84.3%—Not comparable
DeepSearchQASource 93.1%—Not comparable
OSWorld-VerifiedSource 83.4%64.7%Claude Opus 4.8 leads
Finance Agent v2Source 53.9%—Not comparable
GDPval-AASource 1600—Not comparable
MCP AtlasSource 82.2%—Not comparable
ToolathlonSource 59.9%—Not comparable
Gert LabsSource 72.97%57.47%Claude Opus 4.8 leads
AA Agentic IndexSource 47.2%—Not comparable
Tau2-TelecomSource 94.4%86%Claude Opus 4.8 leads
GDPval-AASource 55.0%—Not comparable
ResearchClawBenchSource 21.1%—Not comparable
OSWorld 2.0Source 20.6%—Not comparable
AA BriefcaseSource 1354—Not comparable
AA AutomationBenchSource 48.5%—Not comparable
AA EnterpriseOps-GymSource 44.0%—Not comparable
AA Harvey LABSource 7.5%—Not comparable
AA Tau3 BankingSource 27.6%—Not comparable
JobBenchSource —33.7%Not comparable
CodingClaude Opus 4.8 wins
BenchmarkClaude Opus 4.8GPT-5.3 CodexResult
SWE-bench VerifiedSource 88.6%85%Claude Opus 4.8 leads
SWE-bench ProSource 69.2%56.8%Claude Opus 4.8 leads
SWE MultilingualSource 84.4%—Not comparable
SWE MultimodalSource 38.4%—Not comparable
Terminal-Bench 2.0Source 74.6%—Not comparable
cursorBench31Source 58.4%—Not comparable
cursorBench32Source 62.3%—Not comparable
AA Coding IndexSource 74.3%—Not comparable
Terminal-Bench HardSource 58.3%53.0%Claude Opus 4.8 leads
AA-SciCodeSource 53.5%53.2%Claude Opus 4.8 leads
FrontierCodeSource 46.5%—Not comparable
AA Terminal-Bench 2.1Source 84.6%—Not comparable
SWE-RebenchSource —58.2%Not comparable
Vibe Code BenchSource —61.77%Not comparable
Reasoning
BenchmarkClaude Opus 4.8GPT-5.3 CodexResult
AA-LCRSource 67.7%74.0%GPT-5.3 Codex leads
CritPtSource 20.9%16.9%Claude Opus 4.8 leads
Knowledge
BenchmarkClaude Opus 4.8GPT-5.3 CodexResult
GPQASource 93.6%—Not comparable
GPQA-DSource 93.6%—Not comparable
HLESource 57.9%—Not comparable
HLE w/o toolsSource 49.8%—Not comparable
Artificial Analysis Intelligence IndexSource 55.7%44.3%Claude Opus 4.8 leads
AA-GPQA DiamondSource 92.0%91.5%Claude Opus 4.8 leads
AA-HLESource 45.7%39.9%Claude Opus 4.8 leads
AA-Omniscience IndexSource 27.4%9.9%Claude Opus 4.8 leads
AA-Omniscience AccuracySource 46.6%51.8%GPT-5.3 Codex leads
AA-Omniscience Hallucination RateSource 35.9%86.9%Claude Opus 4.8 leads
Math
BenchmarkClaude Opus 4.8GPT-5.3 CodexResult
USAMO 2026Source 96.7%—Not comparable
FrontierMath v2 (Tiers 1-3)Source 47.241%—Not comparable
FrontierMath v2 (Tier 4)Source 31.250%—Not comparable
Multilingual
BenchmarkClaude Opus 4.8GPT-5.3 CodexResult
INCLUDESource 87.6%—Not comparable
Multimodal
BenchmarkClaude Opus 4.8GPT-5.3 CodexResult
OfficeQA ProSource 66.2%—Not comparable
ScreenSpot ProSource 87.9%—Not comparable
CharXivSource 89.9%—Not comparable
CharXiv w/o toolsSource 80.5%—Not comparable
Design Arena WebsiteSource 12811205Claude Opus 4.8 leads
AA-MMMU-ProSource —78.5%Not comparable
Inst. Following
BenchmarkClaude Opus 4.8GPT-5.3 CodexResult
AA-IFBenchSource 62.2%75.4%GPT-5.3 Codex leads
Frequently Asked Questions (3)

Which is better, Claude Opus 4.8 or GPT-5.3 Codex?

Claude Opus 4.8 is ahead on BenchLM's provisional leaderboard, 85 to 82. The biggest single separator in this matchup is OSWorld-Verified, where the scores are 83.4% and 64.7%.

Which is better for coding, Claude Opus 4.8 or GPT-5.3 Codex?

Claude Opus 4.8 has the edge for coding in this comparison, averaging 76.4 versus 64.4. Inside this category, SWE-bench Pro is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, Claude Opus 4.8 or GPT-5.3 Codex?

Claude Opus 4.8 has the edge for agentic tasks in this comparison, averaging 80.3 versus 71.4. Inside this category, OSWorld-Verified is the benchmark that creates the most daylight between them.

Related Comparisons

Explore More

Anthropic Compare Pricing Methodology Find Your Best LLM Overall Rankings
Last updated: July 12, 2026

The AI models change fast. We track them for you.

A weekly brief for engineers and researchers covering new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.