Kimi K3 vs Muse Spark 1.1
Wins 6 of 8 areas
Coding · Agents · Reasoning · Images and charts · Long documents · Following instructions
Wins 2 of 8 areas
Facts · Math
Kimi K3 is the stronger all-rounder.Muse Spark 1.1 is cheaper and better at facts.
Scores updated · 35 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. 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 software40Kimi K34 of 5 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own41Kimi K34 of 5 tests
- ReasoningHard problems that need careful thinking30Kimi K33 of 4 tests · 1 tie
- Images and chartsUnderstanding pictures, charts and video10Kimi K31 of 1 test
- Long documentsFinding answers in very long texts10Kimi K31 of 1 test
- Following instructionsDoing exactly what it is asked10Kimi K31 of 1 test
- FactsGetting facts right instead of making them up03Muse Spark 1.13 of 3 tests
- MathCompetition and research-level math01Muse Spark 1.11 of 1 test
Images and charts, math, 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.
Muse Spark 1.1 costs 69% 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 K3 pulls ahead
- Real work tasks from 44 professionsGDPVal+16points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+14.2points ahead
- Reasons across sets of long documentsAA-LCR+11points ahead
Where Muse Spark 1.1 pulls ahead
- Short factual questions, answered correctlySimpleQA Verified+7.2points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+4.5points ahead
- Multi-step tasks using many real software toolsMCP Atlas+3.9points ahead
Every test, side by side
All 35 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingKimi K3
- LMArena · WebDevKimi K3 by 113 rating points16551542+113 rating
- Terminal-Bench 2.1Kimi K3 by 7.18577.9+7.1
- LiveBench · CodingKimi K3 by 4.381.577.2+4.3
- LiveBench · Agentic CodingKimi K3 by 3.762.258.5+3.7
- SciCodetie59.558.8tie
AgentsKimi K3
- GDPValKimi K3 by 1651.735.7+16
- τ-Bench V3 · BankingKimi K3 by 14.24631.8+14.2
- Terminal-Bench 4.0Kimi K3 by 6.512.66.1+6.5
- OSWorld-VerifiedKimi K3 by 484.880.8+4
- MCP AtlasMuse Spark 1.1 by 3.984.288.1+3.9
ReasoningKimi K3
- CritPtKimi K3 by 8.323.415.1+8.3
- GPQA DiamondKimi K3 by 3.793.589.8+3.7
- LiveBench · ReasoningKimi K3 by 390.787.7+3
- Humanity's Last Examtie46.946.2tie
FactsMuse Spark 1.1
- SimpleQA VerifiedMuse Spark 1.1 by 7.250.657.8+7.2
- AA-Omniscience · AccuracyMuse Spark 1.1 by 4.547.652+4.5
- AA-Omniscience · Non-hallucinationMuse Spark 1.1 by 3.246.850+3.2
Following instructionsKimi K3
- LiveBench · Instruction FollowingKimi K3 by 1.871.469.6+1.8
Other results14 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.
- GDPval-AA v2 EloKimi K3 by 305 rating points16861381+305 rating
- AA Agentic IndexKimi K3 by 23.150.627.5+23.1
- CyberGymKimi K3 by 218059+21
- DeepSWE 1.1Kimi K3 by 166953+16
- livebench_languageKimi K3 by 11.285.574.3+11.2
- DeepSearchQA (F1)Kimi K3 by 10.19584.9+10.1
- AA IntelligenceKimi K3 by 9.943.633.7+9.9
- BabyVisionKimi K3 by 9.485.776.3+9.4
- AA-OmniscienceMuse Spark 1.1 by 8.419.728.1+8.4
- livebench_data_analysisKimi K3 by 6.178.772.5+6.1
- Artificial Analysis Coding IndexKimi K3 by 4.976.271.3+4.9
- ToolathlonMuse Spark 1.1 by 2.473.275.6+2.4
- JobBenchMuse Spark 1.1 by 1.852.954.7+1.8
- Toolathlon Verifiedtie76.575.6tie
Questions people ask
Which is better, Kimi K3 or Muse Spark 1.1?
Kimi K3 wins six of the eight areas we test: coding, agents, reasoning, images and charts, long documents and following instructions. Muse Spark 1.1 wins facts and math, and costs 69% less.
Which is better for coding?
Kimi K3. It wins 4 of the 5 coding tests both models report; Muse Spark 1.1 wins none, and 1 is a tie.
Which is cheaper?
Kimi K3 costs $3.00 per million input tokens and $15.00 per million output tokens; Muse Spark 1.1 costs $1.25 and $4.25. That makes Muse Spark 1.1 about 69% cheaper for the same work.
How do you compare the two?
We use the 35 benchmark tests both models have published scores on. The verdict counts the 21 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 14 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.