Claude Haiku 4.5 vs Nova 2.0 Pro Preview
Wins 3 of 7 areas
Coding · Agents · Long documents
Wins 2 of 7 areas
Images and charts · Following instructions
Claude Haiku 4.5 wins more areas, narrowly.Nova 2.0 Pro Preview is better at images and charts.
Scores updated · 18 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 software20Claude Haiku 4.52 of 3 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own20Claude Haiku 4.52 of 2 tests
- Long documentsFinding answers in very long texts10Claude Haiku 4.51 of 1 test
- Images and chartsUnderstanding pictures, charts and video01Nova 2.0 Pro Preview1 of 1 test
- Following instructionsDoing exactly what it is asked01Nova 2.0 Pro Preview1 of 1 test
- ReasoningHard problems that need careful thinking11Even1 each · 1 tie
- FactsGetting facts right instead of making them up11Even1 each
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.
Claude Haiku 4.5 costs 47% 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 Haiku 4.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+63.7points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+14.6points ahead
- Reasons across sets of long documentsAA-LCR+10.3points ahead
Where Nova 2.0 Pro Preview pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+24.7points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+11.3points ahead
- Harder college exam questions with imagesMMMU-Pro+5.9points ahead
Every test, side by side
All 18 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
- Terminal-Bench HardClaude Haiku 4.5 by 3.127.324.2+3.1
- SciCodetie42.242.7tie
AgentsClaude Haiku 4.5
- GDPValClaude Haiku 4.5 by 11.811.80+11.8
- τ-Bench V3 · BankingClaude Haiku 4.5 by 1.79.37.6+1.7
ReasoningEven
- GPQA DiamondNova 2.0 Pro Preview by 11.367.278.5+11.3
- Humanity's Last ExamClaude Haiku 4.5 by 110.49.4+1
- CritPttie00tie
FactsEven
- AA-Omniscience · Non-hallucinationClaude Haiku 4.5 by 63.772.79+63.7
- AA-Omniscience · AccuracyNova 2.0 Pro Preview by 41822+4
Images and chartsNova 2.0 Pro Preview
- MMMU-ProNova 2.0 Pro Preview by 5.958.664.5+5.9
Following instructionsNova 2.0 Pro Preview
- IFBenchNova 2.0 Pro Preview by 24.754.379+24.7
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 44.6-4.4-49+44.6
- τ²-Bench Telecom (AA run)Nova 2.0 Pro Preview by 3854.792.7+38
- Artificial Analysis Coding IndexClaude Haiku 4.5 by 9.943.934+9.9
- AA Agentic IndexClaude Haiku 4.5 by 3.310.37+3.3
- AA IntelligenceClaude Haiku 4.5 by 2.716.914.2+2.7
Questions people ask
Which is better, Claude Haiku 4.5 or Nova 2.0 Pro Preview?
Claude Haiku 4.5 wins three of the seven areas where both have results: coding, agents and long documents. Nova 2.0 Pro Preview wins images and charts and following instructions. They are level on reasoning and facts.
Which is better for coding?
Claude Haiku 4.5. It wins 2 of the 3 coding tests both models report; Nova 2.0 Pro Preview wins none, and 1 is a tie.
Which is cheaper?
Claude Haiku 4.5 costs $1.00 per million input tokens and $5.00 per million output tokens; Nova 2.0 Pro Preview costs $1.25 and $10.00. That makes Claude Haiku 4.5 about 47% cheaper for the same work.
How do you compare the two?
We use the 18 benchmark tests both models have published scores on. The verdict counts the 13 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.