Claude Fable 5 vs Claude Opus 4.8
Wins 8 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Long documents · Following instructions
Wins 0 of 8 areas
—
Claude Fable 5 is the stronger all-rounder.Claude Opus 4.8 is cheaper.
Both rank among the ten best models we track in coding and agents.
Scores updated · 75 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 9 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own70Claude Fable 57 of 7 tests
- ReasoningHard problems that need careful thinking40Claude Fable 54 of 6 tests · 2 ties
- MathCompetition and research-level math30Claude Fable 53 of 3 tests
- Following instructionsDoing exactly what it is asked20Claude Fable 52 of 2 tests
- FactsGetting facts right instead of making them up21Claude Fable 52 of 3 tests
- Images and chartsUnderstanding pictures, charts and video10Claude Fable 51 of 1 test
- Long documentsFinding answers in very long texts10Claude Fable 51 of 1 test
Images and charts and long documents rest on a single test each.
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.
Claude Opus 4.8 costs 50% 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+34.1points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+20.7points ahead
- Short factual questions, answered correctlySimpleQA Verified+17.7points ahead
Where Claude Opus 4.8 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+24.3points ahead
Every test, side by side
All 75 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Fable 5
- LiveBench · Agentic CodingClaude Fable 5 by 11.762.250.5+11.7
- SWE-bench ProClaude Fable 5 by 10.88069.2+10.8
- LMArena · WebDevClaude Fable 5 by 68 rating points16241556+68 rating
- SciCodeClaude Fable 5 by 6.66154.4+6.6
- SWE-bench VerifiedClaude Fable 5 by 6.49588.6+6.4
- Terminal-Bench HardClaude Fable 5 by 4.662.958.3+4.6
- LiveBench · CodingClaude Fable 5 by 4.28681.8+4.2
- SWE-bench MultilingualClaude Fable 5 by 2.286.684.4+2.2
- Terminal-Bench 2.1tie84.684.6tie
AgentsClaude Fable 5
- Terminal-Bench 4.0Claude Fable 5 by 20.742.421.7+20.7
- GDPValClaude Fable 5 by 7.855.647.8+7.8
- AA ApexAgentsClaude Fable 5 by 3.943.339.4+3.9
- τ-Bench V3 · BankingClaude Fable 5 by 3.938.134.2+3.9
- BrowseCompClaude Fable 5 by 3.187.484.3+3.1
- MCP AtlasClaude Fable 5 by 2.584.782.2+2.5
- OSWorld-VerifiedClaude Fable 5 by 1.68583.4+1.6
ReasoningClaude Fable 5
- ARC-AGI-2Claude Fable 5 by 17.189.272.1+17.1
- SimpleBenchClaude Fable 5 by 17.181.964.8+17.1
- CritPtClaude Fable 5 by 7.728.620.9+7.7
- Humanity's Last ExamClaude Fable 5 by 6.855.548.7+6.8
- GPQA Diamondtie92.692tie
- LiveBench · Reasoningtie89.789.2tie
FactsClaude Fable 5
- AA-Omniscience · Non-hallucinationClaude Opus 4.8 by 24.336.460.7+24.3
- SimpleQA VerifiedClaude Fable 5 by 17.770.753+17.7
- AA-Omniscience · AccuracyClaude Fable 5 by 16.665.348.8+16.6
Images and chartsClaude Fable 5
- LMArena · VisionClaude Fable 5 by 30 rating points13251295+30 rating
MathClaude Fable 5
- FrontierMath Tier 4Claude Fable 5 by 34.190.256.1+34.1
- FrontierMath Tiers 1-3 (v2)Claude Fable 5 by 78780+7
- LiveBench · MathematicsClaude Fable 5 by 1.79694.3+1.7
Following instructionsClaude Fable 5
- LiveBench · Instruction FollowingClaude Fable 5 by 3.875.872+3.8
- IFBenchClaude Fable 5 by 1.363.562.2+1.3
Other results43 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.
- GDP (Surge AI)Claude Opus 4.8 by 5529.884.8+55
- Blueprint-Bench 2Claude Fable 5 by 24.138.614.5+24.1
- HiL-Bench (Tools-allowed)Claude Fable 5 by 2156.335.3+21
- Terminal-Bench 4.0Claude Fable 5 by 20.944.523.6+20.9
- SWE-bench MultimodalClaude Fable 5 by 15.754.138.4+15.7
- AndroidBenchClaude Fable 5 by 14.784.569.8+14.7
- AA-OmniscienceClaude Fable 5 by 14.543.328.8+14.5
- livebench_data_analysisClaude Fable 5 by 14.580.566+14.5
- Program BenchClaude Fable 5 by 14.486.371.9+14.4
- GDPval-AA v2Claude Fable 5 by 130 rating points17231593+130 rating
- GDPval-AA v2 EloClaude Fable 5 by 130 rating points17231593+130 rating
- DeepSWEClaude Fable 5 by 127058+12
- DeepSWE 1.1Claude Fable 5 by 117059+11
- livebench_languageClaude Fable 5 by 1190.779.7+11
- MCP Mark VerifiedClaude Fable 5 by 1187.476.4+11
- OSWorld 2.0Claude Fable 5 by 10.466.155.7+10.4
- HealthBench ProfessionalClaude Fable 5 by 10.26655.8+10.2
- JobBenchClaude Fable 5 by 957.448.4+9
- SWE-MarathonClaude Fable 5 by 93526+9
- HealthBench Professional (raw)Claude Fable 5 by 8.668.960.3+8.6
- PaperBenchClaude Fable 5 by 8.588.880.3+8.5
- AA Agentic IndexClaude Fable 5 by 8.45142.6+8.4
- OfficeQA ProClaude Opus 4.8 by 8.357.966.2+8.3
- WideSearchClaude Fable 5 by 8.381.272.9+8.3
- AA IntelligenceClaude Fable 5 by 7.849.641.8+7.8
- Toolathlon Pass@∞Claude Fable 5 by 7.555.648.1+7.5
- MLS Bench Litelower is betterClaude Opus 4.8 by 7.149.942.8+7.1
- ARC-AGI-1Claude Fable 5 by 698.592.5+6
- HLE (with tools)Claude Fable 5 by 5.963.857.9+5.9
- Toolathlon Avg turnsClaude Opus 4.8 by 4.719.824.5+4.7
- GDPval-AA (Elo)Claude Fable 5 by 42 rating points19321890+42 rating
- PostTrainBenchClaude Fable 5 by 4.241.437.2+4.2
- τ²-Bench Telecom (AA run)Claude Fable 5 by 4.198.594.4+4.1
- Legal Agent BenchmarkClaude Fable 5 by 3.313.310+3.3
- SpreadsheetBench 2Claude Fable 5 by 3.134.731.6+3.1
- Finance Agent v2Claude Fable 5 by 2.456.353.9+2.4
- HealthBench (raw)Claude Fable 5 by 2.461.258.8+2.4
- Artificial Analysis Coding IndexClaude Fable 5 by 2.276.574.3+2.2
- Toolathlon VerifiedClaude Opus 4.8 by 277.979.9+2
- ToolathlonClaude Fable 5 by 1.861.759.9+1.8
- DeepSearchQA (F1)Claude Fable 5 by 1.194.293.1+1.1
- LiveBenchClaude Fable 5 by 1.178.377.2+1.1
- Agents' Last Examtie25.725.7tie
Questions people ask
Which is better, Claude Fable 5 or Claude Opus 4.8?
Claude Fable 5 wins all eight areas we test: coding, agents, reasoning, facts, images and charts, math, long documents and following instructions. Claude Opus 4.8 wins none, but costs 50% less.
Which is better for coding?
Claude Fable 5. It wins 8 of the 9 coding tests both models report; Claude Opus 4.8 wins none, and 1 is a tie.
Which is cheaper?
Claude Fable 5 costs $10.00 per million input tokens and $50.00 per million output tokens; Claude Opus 4.8 costs $5.00 and $25.00. That makes Claude Opus 4.8 about 50% cheaper for the same work.
How do you compare the two?
We use the 75 benchmark tests both models have published scores on. The verdict counts the 32 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 43 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.