Claude Sonnet 4.6 vs JT-4.1 Flash 236B A21B
Wins 2 of 6 areas
Images and charts · Long documents
Wins 1 of 6 areas
Reasoning
Claude Sonnet 4.6 wins more areas, narrowly.JT-4.1 Flash 236B A21B is better at reasoning.
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.
- Images and chartsUnderstanding pictures, charts and video10Claude Sonnet 4.61 of 1 test
- Long documentsFinding answers in very long texts10Claude Sonnet 4.61 of 1 test
- ReasoningHard problems that need careful thinking12JT-4.1 Flash 236B A21B2 of 3 tests
- CodingWriting and fixing software11Even1 each
- AgentsCarrying out multi-step tasks on its own11Even1 each · 1 tie
- 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 three tests each model wins by the widest margin. Scores are out of 100.
Where Claude Sonnet 4.6 pulls ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+11.6points ahead
- Harder college exam questions with imagesMMMU-Pro+9.2points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+8.6points ahead
Where JT-4.1 Flash 236B A21B pulls ahead
- Research-level physics problemsCritPt+10.9points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+10points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+8.5points 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.
CodingEven
- Terminal-Bench 2.1Claude Sonnet 4.6 by 11.671.259.6+11.6
- SciCodeJT-4.1 Flash 236B A21B by 1.950.152+1.9
AgentsEven
- τ-Bench V3 · BankingClaude Sonnet 4.6 by 6.434.428+6.4
- GDPValJT-4.1 Flash 236B A21B by 4.736.741.4+4.7
- Terminal-Bench 4.0tie33tie
ReasoningJT-4.1 Flash 236B A21B
- CritPtJT-4.1 Flash 236B A21B by 10.93.114+10.9
- Humanity's Last ExamJT-4.1 Flash 236B A21B by 1033.643.6+10
- GPQA DiamondClaude Sonnet 4.6 by 387.584.5+3
FactsEven
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.6 by 8.651.643+8.6
- AA-Omniscience · AccuracyJT-4.1 Flash 236B A21B by 8.540.949.4+8.5
Images and chartsClaude Sonnet 4.6
- MMMU-ProClaude Sonnet 4.6 by 9.273.364.1+9.2
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.
- Artificial Analysis Coding IndexClaude Sonnet 4.6 by 10.66352.4+10.6
- AA-OmniscienceJT-4.1 Flash 236B A21B by 8.412.220.6+8.4
- AA IntelligenceJT-4.1 Flash 236B A21B by 3.830.133.9+3.8
- AA Agentic IndexJT-4.1 Flash 236B A21B by 1.833.134.9+1.8
Questions people ask
Which is better, Claude Sonnet 4.6 or JT-4.1 Flash 236B A21B?
Claude Sonnet 4.6 wins two of the six areas where both have results: images and charts and long documents. JT-4.1 Flash 236B A21B wins reasoning. They are level on coding, agents and facts.
Which is better for coding?
Neither. They win 1 coding test each of the 2 both models report.
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.