Claude Fable 5 vs GPT-5.4
Wins 6 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math
Wins 0 of 8 areas
—
Claude Fable 5 is the stronger all-rounder.GPT-5.4 is cheaper.
Scores updated · 48 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software80Claude Fable 58 of 8 tests
- AgentsCarrying out multi-step tasks on its own51Claude Fable 55 of 6 tests
- ReasoningHard problems that need careful thinking40Claude Fable 54 of 5 tests · 1 tie
- FactsGetting facts right instead of making them up30Claude Fable 53 of 3 tests
- MathCompetition and research-level math30Claude Fable 53 of 3 tests
- Images and chartsUnderstanding pictures, charts and video20Claude Fable 52 of 2 tests
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
- Following instructionsDoing exactly what it is asked11Even1 each
Long documents rests on a single test.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
GPT-5.4 costs 71% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Claude Fable 5 pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+41.2points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+28.1points ahead
- Short factual questions, answered correctlySimpleQA Verified+25.6points ahead
Where GPT-5.4 pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+10.4points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+1.5points ahead
Every test, side by side
All 48 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Fable 5
- SWE-bench ProClaude Fable 5 by 22.38057.7+22.3
- LMArena · WebDevClaude Fable 5 by 160 rating points16251465+160 rating
- SWE-bench MultilingualClaude Fable 5 by 14.986.671.7+14.9
- LiveBench · CodingClaude Fable 5 by 8.58677.5+8.5
- LiveBench · Agentic CodingClaude Fable 5 by 8.462.253.8+8.4
- Terminal-Bench 2.1Claude Fable 5 by 6.384.678.3+6.3
- Terminal-Bench HardClaude Fable 5 by 5.362.957.6+5.3
- SciCodeClaude Fable 5 by 4.46156.6+4.4
AgentsClaude Fable 5
- GDPValClaude Fable 5 by 18.255.637.4+18.2
- MCP AtlasClaude Fable 5 by 17.584.767.2+17.5
- AA ApexAgentsClaude Fable 5 by 1043.333.3+10
- OSWorld-VerifiedClaude Fable 5 by 108575+10
- BrowseCompClaude Fable 5 by 4.787.482.7+4.7
- τ-Bench V3 · BankingGPT-5.4 by 1.538.139.6+1.5
ReasoningClaude Fable 5
- ARC-AGI-2Claude Fable 5 by 15.989.273.3+15.9
- Humanity's Last ExamClaude Fable 5 by 11.855.543.7+11.8
- CritPtClaude Fable 5 by 5.228.623.4+5.2
- LiveBench · ReasoningClaude Fable 5 by 1.689.788.1+1.6
- GPQA Diamondtie92.692tie
FactsClaude Fable 5
- AA-Omniscience · Non-hallucinationClaude Fable 5 by 28.136.48.3+28.1
- SimpleQA VerifiedClaude Fable 5 by 25.670.745.1+25.6
- AA-Omniscience · AccuracyClaude Fable 5 by 14.565.350.9+14.5
Images and chartsClaude Fable 5
- MMMU-ProClaude Fable 5 by 5.884.278.4+5.8
- LMArena · VisionClaude Fable 5 by 23 rating points13251302+23 rating
MathClaude Fable 5
- FrontierMath Tier 4Claude Fable 5 by 41.290.249+41.2
- FrontierMath Tiers 1-3 (v2)Claude Fable 5 by 8.48778.6+8.4
- LiveBench · MathematicsClaude Fable 5 by 1.89694.2+1.8
Following instructionsEven
- IFBenchGPT-5.4 by 10.463.573.9+10.4
- LiveBench · Instruction FollowingClaude Fable 5 by 5.675.870.2+5.6
Other results18 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- HiL-Bench (Tools-allowed)Claude Fable 5 by 46.656.39.7+46.6
- AA-OmniscienceClaude Fable 5 by 37.543.35.8+37.5
- GDPval-AA (Elo)Claude Fable 5 by 258 rating points19321674+258 rating
- DeepSWE 1.1Claude Fable 5 by 187052+18
- DeepSearchQA (F1)Claude Fable 5 by 15.694.278.6+15.6
- Legal Agent BenchmarkClaude Fable 5 by 12.913.30.4+12.9
- HLE (with tools)Claude Fable 5 by 11.863.852+11.8
- τ²-Bench Telecom (AA run)Claude Fable 5 by 11.498.587.1+11.4
- AA IntelligenceClaude Fable 5 by 10.649.639+10.6
- livebench_languageClaude Fable 5 by 8.190.782.6+8.1
- ToolathlonClaude Fable 5 by 7.161.754.6+7.1
- AA Agentic IndexClaude Fable 5 by 6.85144.2+6.8
- OfficeQA ProClaude Fable 5 by 6.857.951.1+6.8
- Artificial Analysis Coding IndexClaude Fable 5 by 5.476.571.1+5.4
- ARC-AGI-1Claude Fable 5 by 4.898.593.7+4.8
- LiveBenchGPT-5.4 by 278.380.3+2
- livebench_data_analysisClaude Fable 5 by 1.280.579.3+1.2
- charxiv_rq_with_pythontie89.490tie
Questions people ask
Which is better, Claude Fable 5 or GPT-5.4?
Claude Fable 5 wins six of the eight areas we test: coding, agents, reasoning, facts, images and charts and math. GPT-5.4 wins none, but costs 71% less. They are level on long documents and following instructions.
Which is better for coding?
Claude Fable 5. It wins 8 of the 8 coding tests both models report; GPT-5.4 wins none.
Which is cheaper?
Claude Fable 5 costs $10.00 per million input tokens and $50.00 per million output tokens; GPT-5.4 costs $2.50 and $15.00. That makes GPT-5.4 about 71% cheaper for the same work.
How do you compare the two?
We use the 48 benchmark tests both models have published scores on. The verdict counts the 30 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 18 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.