Kimi K2 Instruct vs Llama 3.1 Nemotron Ultra 253B V1
Wins 4 of 6 areas
Agents · Facts · Long documents · Following instructions
Wins 0 of 6 areas
—
Kimi K2 Instruct is the stronger all-rounder.
Scores updated · 19 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.
- FactsGetting facts right instead of making them up20Kimi K2 Instruct2 of 2 tests
- AgentsCarrying out multi-step tasks on its own10Kimi K2 Instruct1 of 1 test
- Long documentsFinding answers in very long texts10Kimi K2 Instruct1 of 1 test
- Following instructionsDoing exactly what it is asked10Kimi K2 Instruct1 of 1 test
- CodingWriting and fixing software11Even1 each
- ReasoningHard problems that need careful thinking11Even1 each · 1 tie
Agents, 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 Kimi K2 Instruct pulls ahead
- Reasons across sets of long documentsAA-LCR+45.3points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+21.2points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+11.5points ahead
Where Llama 3.1 Nemotron Ultra 253B V1 pulls ahead
- Code for real scientific research problemsSciCode+4points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+1points ahead
Every test, side by side
All 19 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingEven
- Terminal-Bench HardKimi K2 Instruct by 21.223.52.3+21.2
- SciCodeLlama 3.1 Nemotron Ultra 253B V1 by 430.734.7+4
ReasoningEven
- GPQA DiamondKimi K2 Instruct by 3.976.772.8+3.9
- Humanity's Last ExamLlama 3.1 Nemotron Ultra 253B V1 by 16.47.4+1
- CritPttie00tie
FactsKimi K2 Instruct
- AA-Omniscience · Non-hallucinationKimi K2 Instruct by 11.530.318.8+11.5
- AA-Omniscience · AccuracyKimi K2 Instruct by 5.325.420.1+5.3
Following instructionsKimi K2 Instruct
- IFBenchKimi K2 Instruct by 3.541.738.2+3.5
Other results9 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.
- τ²-Bench Telecom (AA run)Kimi K2 Instruct by 6273.411.4+62
- AA Agentic IndexKimi K2 Instruct by 33.937.73.8+33.9
- AIME 2025Llama 3.1 Nemotron Ultra 253B V1 by 2349.572.5+23
- AA-OmniscienceKimi K2 Instruct by 18.3-26.6-44.9+18.3
- Artificial Analysis Coding IndexKimi K2 Instruct by 12.825.913.1+12.8
- LiveCodeBenchLlama 3.1 Nemotron Ultra 253B V1 by 12.653.766.3+12.6
- AA IntelligenceKimi K2 Instruct by 7.815.37.5+7.8
- MATH-500 (EM)tie97.497tie
- IFEvaltie89.889.5tie
Questions people ask
Which is better, Kimi K2 Instruct or Llama 3.1 Nemotron Ultra 253B V1?
Kimi K2 Instruct wins four of the six areas where both have results: agents, facts, long documents and following instructions. Llama 3.1 Nemotron Ultra 253B V1 wins none. They are level on coding and reasoning.
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 19 benchmark tests both models have published scores on. The verdict counts the 10 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 9 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.