Skip to main content
BenchLM
Advertise

Model comparison

Claude Fable 5 vs GPT-5.4 mini

Data verified

Head-to-head evidence from 18 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

83.68/100
Margin
26.9pts
← winning
56.77/100
1 category wins1 category wins

Public leaderboard positions: Claude Fable 5 #2 (Supported); GPT-5.4 mini #75 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. Claude Fable 5 and GPT-5.4 mini share 18 comparable benchmark results. 2 of 8 categories are comparable. 16 results are unique to Claude Fable 5; 12 to GPT-5.4 mini.

Updated July 23, 2026
Shared results
18
Claude Fable 5 only
16
GPT-5.4 mini only
12
Comparable categories
2 / 8

Pick Claude Fable 5 if you want the stronger benchmark profile. GPT-5.4 mini only becomes the better choice if multimodal & grounded is the priority or you want the cheaper token bill.

Confidence note. This is a partial-evidence comparison with 18 shared benchmark results across 5 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 Fable 5 is clearly ahead on the BenchAlign aggregate, 83.68 to 56.77. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Claude Fable 5's sharpest advantage is in agentic, where it averages 84.6 against 65.7. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 84.3% to 60%. GPT-5.4 mini does hit back in multimodal & grounded, so the answer changes if that is the part of the workload you care about most.

Claude Fable 5 is also the more expensive model on tokens at $10.00 input / $50.00 output per 1M tokens, versus $0.75 input / $4.50 output per 1M tokens for GPT-5.4 mini. That is roughly 11.1x on output cost alone. Claude Fable 5 gives you the larger context window at 1M+, compared with 400K for GPT-5.4 mini.

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 Fable 5 and GPT-5.4 mini
CategoryClaude Fable 5ΔGPT-5.4 mini
AgenticClaude Fable 584.6Margin← 18.9GPT-5.4 mini65.7
MultimodalClaude Fable 557.9Margin→ 18.7GPT-5.4 mini76.6
CodingClaude Fable 589.2MarginNo overlapGPT-5.4 miniNot measured
KnowledgeClaude Fable 5Not measuredMarginNo overlapGPT-5.4 mini47.8
MathClaude Fable 5Not measuredMarginNo overlapGPT-5.4 mini21.7

Decisive benchmark drivers

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

More
A · Claude Fable 5B · GPT-5.4 mini

  1. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 84.3%B 60%
    Winner: Claude Fable 5Δ 24.3
    Terminal-Bench 2.0: Claude Fable 5 scored 84.3%; GPT-5.4 mini scored 60%. Claude Fable 5 wins this benchmark.

  2. OSWorld-Verified

    Agentic
    Source ↗
    A 85%B 72.1%
    Winner: Claude Fable 5Δ 12.9
    OSWorld-Verified: Claude Fable 5 scored 85%; GPT-5.4 mini scored 72.1%. Claude Fable 5 wins this benchmark.

Operational comparison

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

MetricClaude Fable 5GPT-5.4 miniComparison
Input / output priceUSD per 1M tokensClaude Fable 5$10 input / $50 outputGPT-5.4 mini$0.75 input / $4.5 outputGPT-5.4 mini has the lower combined listed price.
Generation speedtokens per secondClaude Fable 5Not availableGPT-5.4 mini201 tok/sA complete speed comparison is not available.
First-answer latencyseconds to first tokenClaude Fable 5Not availableGPT-5.4 mini3.85 sA complete latency comparison is not available.
Context windowmaximum listed tokensClaude Fable 51M+GPT-5.4 mini400KClaude Fable 5 lists the larger context window.

Benchmark Deep Dive

AgenticClaude Fable 5 wins
BenchmarkClaude Fable 5GPT-5.4 miniResult
Terminal-Bench 2.0Source 84.3%60%Claude Fable 5 leads
OSWorld-VerifiedSource 85%72.1%Claude Fable 5 leads
GDPval-AASource 17481171Claude Fable 5 leads
AA Agentic IndexSource 52.8%30.2%Claude Fable 5 leads
τ²-bench resultsSource 98.5%83.3%Claude Fable 5 leads
GDPval-AASource 62.4%33.6%Claude Fable 5 leads
AA BriefcaseSource 1574—Not comparable
AA AutomationBenchSource 48.6%—Not comparable
AA EnterpriseOps-GymSource 51.1%—Not comparable
AA Harvey LABSource 93.6%—Not comparable
AA Tau3 BankingSource 26.8%—Not comparable
terminalBenchHardSource 62.9%—Not comparable
aaTerminalBench21Source 84.6%—Not comparable
MCP AtlasSource —57.7%Not comparable
ToolathlonSource —42.9%Not comparable
APEX-Agents-AASource —28.2%Not comparable
Coding
BenchmarkClaude Fable 5GPT-5.4 miniResult
SWE-bench VerifiedSource 95%—Not comparable
SWE-bench ProSource 80%—Not comparable
FrontierCode 1.1 MainSource 53.5%27.0%Claude Fable 5 leads
Terminal-Bench 2.0Source 84.3%—Not comparable
cursorBench31Source 70.6%—Not comparable
cursorBench32Source 70.5%—Not comparable
VulcanBench v3Source 87.0%—Not comparable
AA Coding IndexSource 76.5%56.1%Claude Fable 5 leads
AA-SciCodeSource 60.2%49.9%Claude Fable 5 leads
Vibe Code BenchSource —47.97%Not comparable
Reasoning
BenchmarkClaude Fable 5GPT-5.4 miniResult
AA-LCRSource 70.0%69.3%Claude Fable 5 leads
CritPtSource 28.6%10.0%Claude Fable 5 leads
Knowledge
BenchmarkClaude Fable 5GPT-5.4 miniResult
Artificial Analysis Intelligence IndexSource 59.9%40.0%Claude Fable 5 leads
AA-GPQA DiamondSource 92.6%87.5%Claude Fable 5 leads
AA-HLESource 53.3%26.6%Claude Fable 5 leads
AA-Omniscience IndexSource 40.2%-18.7%Claude Fable 5 leads
AA-Omniscience AccuracySource 61.4%37.5%Claude Fable 5 leads
AA-Omniscience Hallucination RateSource 54.9%89.8%Claude Fable 5 leads
GPQASource —88%Not comparable
HLESource —41.5%Not comparable
HLE w/o toolsSource —28.2%Not comparable
Math
BenchmarkClaude Fable 5GPT-5.4 miniResult
FrontierMath v2 (Tiers 1-3)Source —28.280%Not comparable
FrontierMath v2 (Tier 4)Source —2.080%Not comparable
MultimodalGPT-5.4 mini wins
BenchmarkClaude Fable 5GPT-5.4 miniResult
Blueprint-Bench 2Source 38.6%—Not comparable
OfficeQA ProSource 57.9%—Not comparable
Design Arena WebsiteSource 1332—Not comparable
MMMU-ProSource —76.6%Not comparable
MMMU-Pro w/ PythonSource —78%Not comparable
AA-MMMU-ProSource —73.3%Not comparable
Inst. Following
BenchmarkClaude Fable 5GPT-5.4 miniResult
AA-IFBenchSource 63.5%73.3%GPT-5.4 mini leads
Frequently Asked Questions (3)

Which is better, Claude Fable 5 or GPT-5.4 mini?

Claude Fable 5 is ahead on BenchLM's BenchAlign leaderboard, 83.68 to 56.77. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 84.3% and 60%.

Which is better for agentic tasks, Claude Fable 5 or GPT-5.4 mini?

Claude Fable 5 has the edge for agentic tasks in this comparison, averaging 84.6 versus 65.7. Inside this category, GDPval-AA is the benchmark that creates the most daylight between them.

Which is better for multimodal and grounded tasks, Claude Fable 5 or GPT-5.4 mini?

GPT-5.4 mini has the edge for multimodal and grounded tasks in this comparison, averaging 76.6 versus 57.9. Claude Fable 5 stays close enough that the answer can still flip depending on your workload.

Related Comparisons

Explore More

Anthropic Compare Pricing Methodology Find Your Best LLM Overall Rankings
Last updated: July 23, 2026

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.