Claude Haiku 4.5 vs JT-4.1 Flash 236B A21B
Wins 0 of 6 areas
—
Wins 5 of 6 areas
Coding · Agents · Reasoning · Images and charts · Long documents
JT-4.1 Flash 236B A21B is the stronger all-rounder.
Scores updated · 16 tests both models report · How we compare
Where each one wins
Tests won in each of the six 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.
- AgentsCarrying out multi-step tasks on its own03JT-4.1 Flash 236B A21B3 of 3 tests
- ReasoningHard problems that need careful thinking03JT-4.1 Flash 236B A21B3 of 3 tests
- CodingWriting and fixing software02JT-4.1 Flash 236B A21B2 of 2 tests
- Images and chartsUnderstanding pictures, charts and video01JT-4.1 Flash 236B A21B1 of 1 test
- Long documentsFinding answers in very long texts01JT-4.1 Flash 236B A21B1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
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.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where JT-4.1 Flash 236B A21B pulls ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+33.2points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+31.4points ahead
- Real work tasks from 44 professionsGDPVal+29.6points ahead
Where Claude Haiku 4.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+29.7points ahead
Every test, side by side
All 16 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingJT-4.1 Flash 236B A21B
- Terminal-Bench 2.1JT-4.1 Flash 236B A21B by 15.444.259.6+15.4
- SciCodeJT-4.1 Flash 236B A21B by 9.842.252+9.8
AgentsJT-4.1 Flash 236B A21B
- GDPValJT-4.1 Flash 236B A21B by 29.611.841.4+29.6
- τ-Bench V3 · BankingJT-4.1 Flash 236B A21B by 18.79.328+18.7
- Terminal-Bench 4.0JT-4.1 Flash 236B A21B by 303+3
ReasoningJT-4.1 Flash 236B A21B
- Humanity's Last ExamJT-4.1 Flash 236B A21B by 33.210.443.6+33.2
- GPQA DiamondJT-4.1 Flash 236B A21B by 17.367.284.5+17.3
- CritPtJT-4.1 Flash 236B A21B by 14014+14
FactsEven
- AA-Omniscience · AccuracyJT-4.1 Flash 236B A21B by 31.41849.4+31.4
- AA-Omniscience · Non-hallucinationClaude Haiku 4.5 by 29.772.743+29.7
Images and chartsJT-4.1 Flash 236B A21B
- MMMU-ProJT-4.1 Flash 236B A21B by 5.558.664.1+5.5
Long documentsJT-4.1 Flash 236B A21B
- AA-LCRJT-4.1 Flash 236B A21B by 474.378.3+4
Other results4 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-OmniscienceJT-4.1 Flash 236B A21B by 25-4.420.6+25
- AA Agentic IndexJT-4.1 Flash 236B A21B by 24.610.334.9+24.6
- AA IntelligenceJT-4.1 Flash 236B A21B by 1716.933.9+17
- Artificial Analysis Coding IndexJT-4.1 Flash 236B A21B by 8.543.952.4+8.5
Questions people ask
Which is better, Claude Haiku 4.5 or JT-4.1 Flash 236B A21B?
JT-4.1 Flash 236B A21B wins five of the six areas where both have results: coding, agents, reasoning, images and charts and long documents. Claude Haiku 4.5 wins none. They are level on facts.
Which is better for coding?
JT-4.1 Flash 236B A21B. It wins 2 of the 2 coding tests both models report; Claude Haiku 4.5 wins none.
How do you compare the two?
We use the 16 benchmark tests both models have published scores on. The verdict counts the 12 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 4 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.