Claude Fable 5.1 vs Claude Haiku 4.5
Wins 5 of 5 areas
Coding · Agents · Reasoning · Facts · Long documents
Wins 0 of 5 areas
—
Claude Fable 5.1 is the stronger all-rounder.Claude Haiku 4.5 is cheaper.
Scores updated · 40 tests both models report · How we compare
Where each one wins
Tests won in each of the five areas where both have results. 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 software50Claude Fable 5.15 of 5 tests
- AgentsCarrying out multi-step tasks on its own40Claude Fable 5.14 of 4 tests
- ReasoningHard problems that need careful thinking40Claude Fable 5.14 of 4 tests
- FactsGetting facts right instead of making them up21Claude Fable 5.12 of 3 tests
- Long documentsFinding answers in very long texts10Claude Fable 5.11 of 1 test
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.
Claude Haiku 4.5 costs 90% 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 5.1 pulls ahead
- Abstract visual puzzles that people can solveARC-AGI-2+86points ahead
- Short factual questions, answered correctlySimpleQA Verified+57.6points ahead
- Real work tasks from 44 professionsGDPVal+51.1points ahead
Where Claude Haiku 4.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+45.3points ahead
Every test, side by side
All 40 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Fable 5.1
- Terminal-Bench 2.1Claude Fable 5.1 by 47.291.444.2+47.2
- LMArena · WebDevClaude Fable 5.1 by 420 rating points17491329+420 rating
- SWE-bench ProClaude Fable 5.1 by 41.781.239.5+41.7
- SWE-bench MultilingualClaude Fable 5.1 by 21.789.167.4+21.7
- SciCodeClaude Fable 5.1 by 20.963.142.2+20.9
AgentsClaude Fable 5.1
- Terminal-Bench 4.0Claude Fable 5.1 by 52520+52
- GDPValClaude Fable 5.1 by 51.162.911.8+51.1
- τ-Bench V3 · BankingClaude Fable 5.1 by 37.947.29.3+37.9
- AA IT-Bench SREClaude Fable 5.1 by 22.249.527.3+22.2
ReasoningClaude Fable 5.1
- ARC-AGI-2Claude Fable 5.1 by 86904+86
- Humanity's Last ExamClaude Fable 5.1 by 48.759.110.4+48.7
- CritPtClaude Fable 5.1 by 29.729.70+29.7
- GPQA DiamondClaude Fable 5.1 by 26.593.767.2+26.5
FactsClaude Fable 5.1
- SimpleQA VerifiedClaude Fable 5.1 by 57.670.813.2+57.6
- AA-Omniscience · AccuracyClaude Fable 5.1 by 49.267.218+49.2
- AA-Omniscience · Non-hallucinationClaude Haiku 4.5 by 45.327.472.7+45.3
Other results23 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.
- GDPval-AA 2.1Claude Fable 5.1 by 1000 rating points1735735+1000 rating
- Terminal-Bench 4.0Claude Fable 5.1 by 55.855.80+55.8
- ARC-AGI-1Claude Fable 5.1 by 49.897.547.7+49.8
- AA-OmniscienceClaude Fable 5.1 by 47.943.5-4.4+47.9
- AA Agentic IndexClaude Fable 5.1 by 47.75810.3+47.7
- Artificial Analysis Coding IndexClaude Fable 5.1 by 37.781.643.9+37.7
- AA IntelligenceClaude Fable 5.1 by 36.553.416.9+36.5
- SWE-bench MultimodalClaude Fable 5.1 by 34.954.719.8+34.9
- HealthBench Professional length-adjustedClaude Fable 5.1 by 29.962.132.2+29.9
- Suicide and self-harm multi-turn appropriate response rate (Claude.ai)Claude Fable 5.1 by 239471+23
- Suicide and self-harm multi-turn appropriate response rate (API, without a system prompt)Claude Fable 5.1 by 146046+14
- Child safety multi-turn appropriate response rate (API, without a system prompt)Claude Haiku 4.5 by 118495+11
- Election integrity multi-turn (appropriate response rate) - API, without a system promptClaude Haiku 4.5 by 39093+3
- Child safety multi-turn appropriate response rate (Claude.ai)Claude Fable 5.1 by 210098+2
- Election integrity multi-turn (appropriate response rate) - Claude.aiClaude Fable 5.1 by 18887+1
- Election integrity single-turn harmful requests (harmless rate) - API, without a system prompttie10099.8tie
- Election integrity single-turn harmful requests (harmless rate) - Claude.aitie10099.8tie
- Suicide and self-harm single-turn requests posing potential risk harmless rate (API, without a system prompt)tie99.399.5tie
- Suicide and self-harm single-turn requests posing potential risk harmless rate (Claude.ai)tie10099.8tie
- Suicide and self-harm single-turn benign requests refusal rate (Claude.ai)tie0.50.6tie
- Election integrity single-turn benign requests (refusal rate) - API, without a system prompttie00tie
- Election integrity single-turn benign requests (refusal rate) - Claude.aitie00tie
- Suicide and self-harm single-turn benign requests refusal rate (API, without a system prompt)tie00tie
Questions people ask
Which is better, Claude Fable 5.1 or Claude Haiku 4.5?
Claude Fable 5.1 wins all five areas where both have results: coding, agents, reasoning, facts and long documents. Claude Haiku 4.5 wins none, but costs 90% less.
Which is better for coding?
Claude Fable 5.1. It wins 5 of the 5 coding tests both models report; Claude Haiku 4.5 wins none.
Which is cheaper?
Claude Fable 5.1 costs $10.00 per million input tokens and $50.00 per million output tokens; Claude Haiku 4.5 costs $1.00 and $5.00. That makes Claude Haiku 4.5 about 90% cheaper for the same work.
How do you compare the two?
We use the 40 benchmark tests both models have published scores on. The verdict counts the 17 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 23 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.