Claude Fable 5 vs GPT-5.6 Terra
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-5.6 Terra is cheaper.
Both rank among the ten best models we track in coding.
Scores updated · 47 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 software61Claude Fable 56 of 7 tests
- FactsGetting facts right instead of making them up30Claude Fable 53 of 3 tests
- MathCompetition and research-level math30Claude Fable 53 of 3 tests
- ReasoningHard problems that need careful thinking31Claude Fable 53 of 6 tests · 2 ties
- AgentsCarrying out multi-step tasks on its own31Claude Fable 53 of 5 tests · 1 tie
- 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-5.6 Terra costs 77% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Claude Fable 5 pulls ahead
- Common-sense trick questionsSimpleBench+33points ahead
- Short factual questions, answered correctlySimpleQA Verified+27.5points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+24.3points ahead
Where GPT-5.6 Terra pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+7.7points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+3.4points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+2.1points ahead
Every test, side by side
All 47 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 16.68063.4+16.6
- LMArena · WebDevClaude Fable 5 by 102 rating points16241522+102 rating
- LiveBench · CodingClaude Fable 5 by 7.78678.3+7.7
- LiveBench · Agentic CodingClaude Fable 5 by 7.262.255+7.2
- SciCodeClaude Fable 5 by 66155+6
- Terminal-Bench HardClaude Fable 5 by 5.362.957.6+5.3
- Terminal-Bench 2.1GPT-5.6 Terra by 3.484.688+3.4
AgentsClaude Fable 5
- GDPValClaude Fable 5 by 7.955.647.7+7.9
- Terminal-Bench 4.0Claude Fable 5 by 742.435.4+7
- AA ApexAgentsClaude Fable 5 by 4.443.338.9+4.4
- τ-Bench V3 · BankingGPT-5.6 Terra by 2.138.140.2+2.1
- BrowseComptie87.487.5tie
ReasoningClaude Fable 5
- SimpleBenchClaude Fable 5 by 3381.948.9+33
- Humanity's Last ExamClaude Fable 5 by 12.655.542.9+12.6
- ARC-AGI-2Claude Fable 5 by 5.389.283.9+5.3
- CritPtGPT-5.6 Terra by 1.428.630+1.4
- LiveBench · Reasoningtie89.790.6tie
- GPQA Diamondtie92.692.5tie
FactsClaude Fable 5
- SimpleQA VerifiedClaude Fable 5 by 27.570.743.2+27.5
- AA-Omniscience · Non-hallucinationClaude Fable 5 by 24.336.412.1+24.3
- AA-Omniscience · AccuracyClaude Fable 5 by 18.665.346.8+18.6
Images and chartsClaude Fable 5
- LMArena · VisionClaude Fable 5 by 54 rating points13251271+54 rating
- MMMU-ProClaude Fable 5 by 3.584.280.7+3.5
MathClaude Fable 5
- FrontierMath Tier 4Claude Fable 5 by 21.990.268.3+21.9
- LiveBench · MathematicsClaude Fable 5 by 1.19694.9+1.1
- FrontierMath Tiers 1-3 (v2)Claude Fable 5 by 18786+1
Following instructionsEven
- LiveBench · Instruction FollowingClaude Fable 5 by 11.275.864.6+11.2
- IFBenchGPT-5.6 Terra by 7.763.571.2+7.7
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.
- AA-OmniscienceClaude Fable 5 by 43.243.30.1+43.2
- Agents' Last ExamGPT-5.6 Terra by 24.725.750.4+24.7
- Terminal-Bench 4.0Claude Fable 5 by 2344.521.5+23
- OSWorld 2.0 (partial)Claude Fable 5 by 22.772.950.2+22.7
- GDPval-AA v2 EloClaude Fable 5 by 195 rating points17231528+195 rating
- OSWorld 2.0Claude Fable 5 by 15.966.150.2+15.9
- τ²-Bench Telecom (AA run)Claude Fable 5 by 12.298.586.3+12.2
- ToolathlonClaude Fable 5 by 8.661.753.1+8.6
- HealthBench ProfessionalClaude Fable 5 by 8.36657.7+8.3
- livebench_languageClaude Fable 5 by 7.890.782.9+7.8
- AA IntelligenceClaude Fable 5 by 7.549.642.1+7.5
- AA Agentic IndexClaude Fable 5 by 7.35143.7+7.3
- GDP (Surge AI)Claude Fable 5 by 5.129.824.7+5.1
- ARC-AGI-1Claude Fable 5 by 298.596.5+2
- livebench_data_analysisClaude Fable 5 by 1.280.579.3+1.2
- DeepSWEtie7069.6tie
- Artificial Analysis Coding Indextie76.576.7tie
- DeepSWE 1.1tie7070tie
Questions people ask
Which is better, Claude Fable 5 or GPT-5.6 Terra?
Claude Fable 5 wins six of the eight areas we test: coding, agents, reasoning, facts, images and charts and math. GPT-5.6 Terra wins none, but costs 77% less. They are level on long documents and following instructions.
Which is better for coding?
Claude Fable 5. It wins 6 of the 7 coding tests both models report; GPT-5.6 Terra wins 1.
Which is cheaper?
Claude Fable 5 costs $10.00 per million input tokens and $50.00 per million output tokens; GPT-5.6 Terra costs $2.00 and $12.00. That makes GPT-5.6 Terra about 77% cheaper for the same work.
How do you compare the two?
We use the 47 benchmark tests both models have published scores on. The verdict counts the 29 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.