Kimi K2 Instruct vs Qwen3 235B A22B Instruct 2507
Wins 4 of 6 areas
Coding · Agents · Facts · Long documents
Wins 1 of 6 areas
Following instructions
Kimi K2 Instruct is the stronger all-rounder.Qwen3 235B A22B Instruct 2507 is cheaper and better at following instructions.
Scores updated · 33 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
- CodingWriting and fixing software21Kimi K2 Instruct2 of 3 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 asked01Qwen3 235B A22B Instruct 25071 of 1 test
- 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.
Qwen3 235B A22B Instruct 2507 costs 63% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Kimi K2 Instruct pulls ahead
- Reasons across sets of long documentsAA-LCR+19.1points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+8.3points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+7.5points ahead
Where Qwen3 235B A22B Instruct 2507 pulls ahead
- Code for real scientific research problemsSciCode+5.3points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+4.7points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+4.4points ahead
Every test, side by side
All 33 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingKimi K2 Instruct
- Terminal-Bench HardKimi K2 Instruct by 8.323.515.2+8.3
- SciCodeQwen3 235B A22B Instruct 2507 by 5.330.736+5.3
- LiveCodeBench v6Kimi K2 Instruct by 1.953.751.8+1.9
ReasoningEven
- Humanity's Last ExamQwen3 235B A22B Instruct 2507 by 4.76.411.1+4.7
- GPQA DiamondKimi K2 Instruct by 1.476.775.3+1.4
- CritPttie00tie
FactsKimi K2 Instruct
- AA-Omniscience · Non-hallucinationKimi K2 Instruct by 7.530.322.8+7.5
- AA-Omniscience · AccuracyKimi K2 Instruct by 6.725.418.7+6.7
Following instructionsQwen3 235B A22B Instruct 2507
- IFBenchQwen3 235B A22B Instruct 2507 by 4.441.746.1+4.4
Other results22 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 40.173.433.3+40.1
- SimpleQAQwen3 235B A22B Instruct 2507 by 23.33154.3+23.3
- AIME 2025Qwen3 235B A22B Instruct 2507 by 20.849.570.3+20.8
- HarmBenchKimi K2 Instruct by 17.897.479.6+17.8
- AA-OmniscienceKimi K2 Instruct by 17.4-26.6-44+17.4
- HMMT 2025Qwen3 235B A22B Instruct 2507 by 16.638.855.4+16.6
- AA Agentic IndexKimi K2 Instruct by 14.937.722.8+14.9
- Tau2 airlineKimi K2 Instruct by 12.556.544+12.5
- ZebraLogicQwen3 235B A22B Instruct 2507 by 68995+6
- SuperGPQAQwen3 235B A22B Instruct 2507 by 5.457.262.6+5.4
- AIR-Bench 2024Qwen3 235B A22B Instruct 2507 by 4.974.179+4.9
- Artificial Analysis Coding IndexKimi K2 Instruct by 3.825.922.1+3.8
- AA IntelligenceKimi K2 Instruct by 3.315.312+3.3
- Aider-PolyglotKimi K2 Instruct by 2.76057.3+2.7
- MultiPL-EQwen3 235B A22B Instruct 2507 by 2.285.787.9+2.2
- MMLU-ProQwen3 235B A22B Instruct 2507 by 1.981.183+1.9
- IFEvalKimi K2 Instruct by 1.189.888.7+1.1
- τ²-Bench (Retail)tie70.671.3tie
- anthropic_red_teamtie99.399.9tie
- MMLU-Reduxtie92.793.1tie
- XSTesttie98.298.6tie
- simple_safety_teststie100100tie
Questions people ask
Which is better, Kimi K2 Instruct or Qwen3 235B A22B Instruct 2507?
Kimi K2 Instruct wins four of the six areas where both have results: coding, agents, facts and long documents. Qwen3 235B A22B Instruct 2507 wins following instructions, and costs 63% less. They are level on reasoning.
Which is better for coding?
Kimi K2 Instruct. It wins 2 of the 3 coding tests both models report; Qwen3 235B A22B Instruct 2507 wins 1.
Which is cheaper?
Kimi K2 Instruct costs $0.60 per million input tokens and $2.50 per million output tokens; Qwen3 235B A22B Instruct 2507 costs $0.23 and $0.92. That makes Qwen3 235B A22B Instruct 2507 about 63% cheaper for the same work.
How do you compare the two?
We use the 33 benchmark tests both models have published scores on. The verdict counts the 11 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 22 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.