Muse Spark 1.1 wins more areas, narrowly.
Scores updated · 25 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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.
- CodingWriting and fixing software03Muse Spark 1.13 of 4 tests · 1 tie
- ReasoningHard problems that need careful thinking03Muse Spark 1.13 of 3 tests
- AgentsCarrying out multi-step tasks on its own02Muse Spark 1.12 of 2 tests
- FactsGetting facts right instead of making them up12Muse Spark 1.12 of 3 tests
- Images and chartsUnderstanding pictures, charts and video11Even1 each
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
- Following instructionsDoing exactly what it is asked00Even0 each · 1 tie
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 Muse Spark 1.1 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+34.2points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+15.7points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+12.2points ahead
Where Muse Spark pulls ahead
- Short factual questions, answered correctlySimpleQA Verified+8.5points ahead
Every test, side by side
All 25 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingMuse Spark 1.1
- Terminal-Bench 2.1Muse Spark 1.1 by 15.762.277.9+15.7
- SWE-bench ProMuse Spark 1.1 by 9.152.461.5+9.1
- SciCodeMuse Spark 1.1 by 7.351.558.8+7.3
- LMArena · WebDevtie15401542tie
AgentsMuse Spark 1.1
- τ-Bench V3 · BankingMuse Spark 1.1 by 12.219.631.8+12.2
- GDPValMuse Spark 1.1 by 10.625.135.7+10.6
ReasoningMuse Spark 1.1
- Humanity's Last ExamMuse Spark 1.1 by 5.540.746.2+5.5
- CritPtMuse Spark 1.1 by 3.811.315.1+3.8
- GPQA DiamondMuse Spark 1.1 by 1.488.489.8+1.4
FactsMuse Spark 1.1
- AA-Omniscience · Non-hallucinationMuse Spark 1.1 by 34.215.850+34.2
- SimpleQA VerifiedMuse Spark by 8.566.357.8+8.5
- AA-Omniscience · AccuracyMuse Spark 1.1 by 2.549.652+2.5
Images and chartsEven
- CharXiv (RQ)Muse Spark 1.1 by 286.488.4+2
- LMArena · VisionMuse Spark by 11 rating points13051294+11 rating
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.
- HLE CalibrationMuse Spark by 26.950.323.4+26.9
- AA-OmniscienceMuse Spark 1.1 by 20.97.228.1+20.9
- CyberGymMuse Spark 1.1 by 15.543.559+15.5
- Artificial Analysis Coding IndexMuse Spark 1.1 by 12.758.671.3+12.7
- DeepSearchQA (F1)Muse Spark 1.1 by 10.174.884.9+10.1
- OR-Bench (FRR)Muse Spark by 3.284.8+3.2
- AA IntelligenceMuse Spark 1.1 by 2.431.333.7+2.4
- WMDPMuse Spark 1.1 by 1.485.687+1.4
- AA Agentic IndexMuse Spark by 1.228.727.5+1.2
Questions people ask
Which is better, Muse Spark or Muse Spark 1.1?
Muse Spark 1.1 wins four of the seven areas where both have results: coding, agents, reasoning and facts. Muse Spark wins none. They are level on images and charts, long documents and following instructions.
Which is better for coding?
Muse Spark 1.1. It wins 3 of the 4 coding tests both models report; Muse Spark wins none, and 1 is a tie.
How do you compare the two?
We use the 25 benchmark tests both models have published scores on. The verdict counts the 16 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.