Claude Sonnet 4.5 vs o4-mini
Wins 4 of 8 areas
Coding · Reasoning · Facts · Long documents
Wins 3 of 8 areas
Agents · Images and charts · Math
Claude Sonnet 4.5 wins more areas, narrowly.o4-mini is cheaper and better at images and charts.
Scores updated · 42 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.
- FactsGetting facts right instead of making them up40Claude Sonnet 4.54 of 4 tests
- ReasoningHard problems that need careful thinking30Claude Sonnet 4.53 of 4 tests · 1 tie
- CodingWriting and fixing software20Claude Sonnet 4.52 of 3 tests · 1 tie
- Long documentsFinding answers in very long texts10Claude Sonnet 4.51 of 1 test
- Images and chartsUnderstanding pictures, charts and video03o4-mini3 of 4 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own02o4-mini2 of 2 tests
- MathCompetition and research-level math02o4-mini2 of 2 tests
- 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.
o4-mini costs 69% 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 Sonnet 4.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+31.4points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+20.4points ahead
- Keeps track of context across a multi-turn chatMulti-Challenge+12.3points ahead
Where o4-mini pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+27.4points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+12.2points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+11.4points ahead
Every test, side by side
All 42 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 4.5
- Terminal-Bench HardClaude Sonnet 4.5 by 20.435.615.2+20.4
- SWE-bench VerifiedClaude Sonnet 4.5 by 9.177.268.1+9.1
- SciCodetie45.746.5tie
Agentso4-mini
- BrowseCompo4-mini by 27.424.151.5+27.4
- GDPValo4-mini by 4.920.525.4+4.9
ReasoningClaude Sonnet 4.5
- ARC-AGI-2Claude Sonnet 4.5 by 7.513.66.1+7.5
- GPQA DiamondClaude Sonnet 4.5 by 583.478.4+5
- Humanity's Last ExamClaude Sonnet 4.5 by 1.317.816.5+1.3
- CritPttie1.10.6tie
FactsClaude Sonnet 4.5
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.5 by 31.450.919.5+31.4
- SimpleQA VerifiedClaude Sonnet 4.5 by 11.930.718.8+11.9
- AA-Omniscience · AccuracyClaude Sonnet 4.5 by 8.132.924.8+8.1
- Vectara HHEM hallucination ratelower is betterClaude Sonnet 4.5 by 6.61218.6+6.6
Images and chartso4-mini
- CharXiv (RQ)o4-mini by 4.867.272+4.8
- MathVistao4-mini by 4.579.884.3+4.5
- MMMUo4-mini by 279.681.6+2
- MMMU-Protie68.769.2tie
Matho4-mini
- FrontierMath Tiers 1-3 (v2)o4-mini by 12.223.936.1+12.2
- FrontierMath Tier 4o4-mini by 2.52.44.9+2.5
Long documentsClaude Sonnet 4.5
- AA-LCRClaude Sonnet 4.5 by 11.372.361+11.3
Following instructionsEven
- Multi-ChallengeClaude Sonnet 4.5 by 12.355.343+12.3
- IFBencho4-mini by 11.457.368.7+11.4
Other results20 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 Sonnet 4.5 by 35.6-0.1-35.7+35.6
- Artificial Analysis Coding IndexClaude Sonnet 4.5 by 26.552.125.6+26.5
- τ²-Bench Telecom (AA run)Claude Sonnet 4.5 by 22.578.155.6+22.5
- TAU-bench (airline)Claude Sonnet 4.5 by 20.87049.2+20.8
- AA Agentic Indexo4-mini by 18.617.536.1+18.6
- TAU-bench (retail)Claude Sonnet 4.5 by 14.486.271.8+14.4
- AIR-Bench 2024Claude Sonnet 4.5 by 11.389.878.5+11.3
- vectara_factual_consistencyClaude Sonnet 4.5 by 6.68881.4+6.6
- LiveCodeBencho4-mini by 6.47177.4+6.4
- AIME 2025o4-mini by 5.78792.7+5.7
- ARC-AGI-1Claude Sonnet 4.5 by 563.758.7+5
- HarmBencho4-mini by 59297+5
- bbqClaude Sonnet 4.5 by 4.998.994+4.9
- AA IntelligenceClaude Sonnet 4.5 by 420.716.7+4
- vectara_answer_rateo4-mini by 3.695.699.2+3.6
- frontiermath_tier_4_v1o4-mini by 2.14.26.3+2.1
- XSTesto4-mini by 196.497.4+1
- vectara_avg_summary_lengthtie127.8127.7tie
- anthropic_red_teamtie98.298.2tie
- simple_safety_teststie100100tie
Questions people ask
Which is better, Claude Sonnet 4.5 or o4-mini?
Claude Sonnet 4.5 wins four of the eight areas we test: coding, reasoning, facts and long documents. o4-mini wins agents, images and charts and math, and costs 69% less. They are level on following instructions.
Which is better for coding?
Claude Sonnet 4.5. It wins 2 of the 3 coding tests both models report; o4-mini wins none, and 1 is a tie.
Which is cheaper?
Claude Sonnet 4.5 costs $3.00 per million input tokens and $15.00 per million output tokens; o4-mini costs $1.10 and $4.40. That makes o4-mini about 69% cheaper for the same work.
How do you compare the two?
We use the 42 benchmark tests both models have published scores on. The verdict counts the 22 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 20 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.