Claude Haiku 4.5 vs Devstral Small 2
Wins 6 of 7 areas
Coding · Reasoning · Facts · Images and charts · Long documents · Following instructions
Wins 0 of 7 areas
—
Claude Haiku 4.5 is the stronger all-rounder.
Scores updated · 21 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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 Haiku 4.55 of 5 tests
- ReasoningHard problems that need careful thinking20Claude Haiku 4.52 of 3 tests · 1 tie
- FactsGetting facts right instead of making them up20Claude Haiku 4.52 of 2 tests
- Images and chartsUnderstanding pictures, charts and video10Claude Haiku 4.51 of 1 test
- Long documentsFinding answers in very long texts10Claude Haiku 4.51 of 1 test
- Following instructionsDoing exactly what it is asked10Claude Haiku 4.51 of 1 test
- AgentsCarrying out multi-step tasks on its own11Even1 each · 1 tie
Images and charts, long documents and following instructions 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.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Claude Haiku 4.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+60points ahead
- Reasons across sets of long documentsAA-LCR+46.3points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+23.1points ahead
Where Devstral Small 2 pulls ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+1.4points ahead
Every test, side by side
All 21 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Haiku 4.5
- Terminal-Bench 2.1Claude Haiku 4.5 by 14.644.229.6+14.6
- SWE-bench MultilingualClaude Haiku 4.5 by 11.767.455.7+11.7
- Terminal-Bench HardClaude Haiku 4.5 by 10.627.316.7+10.6
- SciCodeClaude Haiku 4.5 by 9.842.232.4+9.8
- SWE-bench VerifiedClaude Haiku 4.5 by 5.373.368+5.3
AgentsEven
- GDPValClaude Haiku 4.5 by 10.211.81.6+10.2
- τ-Bench V3 · BankingDevstral Small 2 by 1.49.310.7+1.4
- Terminal-Bench 4.0tie00tie
ReasoningClaude Haiku 4.5
- GPQA DiamondClaude Haiku 4.5 by 1467.253.2+14
- Humanity's Last ExamClaude Haiku 4.5 by 6.910.43.5+6.9
- CritPttie00tie
FactsClaude Haiku 4.5
- AA-Omniscience · Non-hallucinationClaude Haiku 4.5 by 6072.712.7+60
- AA-Omniscience · AccuracyClaude Haiku 4.5 by 2.21815.8+2.2
Images and chartsClaude Haiku 4.5
- MMMU-ProClaude Haiku 4.5 by 1458.644.6+14
Following instructionsClaude Haiku 4.5
- IFBenchClaude Haiku 4.5 by 23.154.331.2+23.1
Other results5 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 Haiku 4.5 by 53.3-4.4-57.7+53.3
- τ²-Bench Telecom (AA run)Claude Haiku 4.5 by 31.354.723.4+31.3
- Artificial Analysis Coding IndexClaude Haiku 4.5 by 14.643.929.3+14.6
- AA IntelligenceClaude Haiku 4.5 by 9.416.97.5+9.4
- AA Agentic IndexClaude Haiku 4.5 by 5.510.34.8+5.5
Questions people ask
Which is better, Claude Haiku 4.5 or Devstral Small 2?
Claude Haiku 4.5 wins six of the seven areas where both have results: coding, reasoning, facts, images and charts, long documents and following instructions. Devstral Small 2 wins none. They are level on agents.
Which is better for coding?
Claude Haiku 4.5. It wins 5 of the 5 coding tests both models report; Devstral Small 2 wins none.
How do you compare the two?
We use the 21 benchmark tests both models have published scores on. The verdict counts the 16 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 5 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.