Kimi K2.7 Code vs Muse Spark Open
Full profiles: Kimi K2.7 Code (Moonshot AI) and Muse Spark Open (Meta): every benchmark, price and rank.
Updated · Data from 12 sources
As of 16 Nov 2026, Kimi K2.7 Code and Muse Spark Open are neck and neck on quality (53.9 vs 53.5 on the BenchLeader Index, inside the ±8.0 margin), and Kimi K2.7 Code is 1.2× cheaper at $1.71 vs $2.00 per 1M tokens blended. Choose Kimi K2.7 Code for coding (52 vs 36) and reasoning (53 vs 51); choose Muse Spark Open for fast first answers (9.9 s vs 29.6 s) and knowledge (59 vs 45), where it also streams 3×+ faster.
Key facts (as of 16 Nov 2026)
- Kimi K2.7 Code scores 53.9 on the BenchLeader Index and Muse Spark Open scores 53.5 (BenchLeader, 16 Nov 2026).
- Kimi K2.7 Code costs $1.71 per 1M tokens blended (list price, 16 Nov 2026).
- Muse Spark Open costs $2.00 per 1M tokens blended (list price, 16 Nov 2026).
- Muse Spark Open streams 3×+ faster than Kimi K2.7 Code (Artificial Analysis, 7-day median, 16 Nov 2026).
- Of the 19 benchmarks both models have, Muse Spark Open leads 14 (BenchLeader, 16 Nov 2026).
Which to pick for…
The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.
| Use case | Pick | Why | Confidence |
|---|---|---|---|
| Coding | Kimi K2.7 Code | 52 vs 36 category score; Code Migration 25.4% vs 21.2% | High |
| Reasoning | Kimi K2.7 Code | 53 vs 51 category score; HLE-Diamond (AA) 14.8% vs 26.3% | Medium |
| Knowledge | Muse Spark Open | 59 vs 45 category score | High |
| Long documents | Either | 64 vs 64 category score; AA-LCR v1.1 79.3% vs 78.3% | Either |
| Lowest cost | Kimi K2.7 Code | $1.71 vs $2.00 per 1M blended; agentic coding session $0.073 vs $0.092 | – |
| Fastest first answer | Muse Spark Open | 9.9 s (Muse Spark Open (xhigh)) vs 29.6 s (Kimi K2.7 Code) to the first answer token | – |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- Kimi K2.7 Code and Muse Spark Open (xhigh) are neck and neck on quality (53.9 vs 53.5, within the ±8.0 margin).
- Kimi K2.7 Code is stronger in coding, long context, reasoning.
- Muse Spark Open (xhigh) is stronger in knowledge.
- Kimi K2.7 Code is 1.2× cheaper ($1.71 vs $2.00 per 1M blended).
- Muse Spark Open (xhigh) streams 3.3× faster (180 vs 55 tokens per second).
At a glance
| Spec | Kimi K2.7 Code | Muse Spark Open |
|---|---|---|
| Developer | Moonshot AI | Meta |
| Released | 12 Jun 2026 | 15 Oct 2026 |
| Context window (tokens) | 262,144 | 1,048,576 |
| Compared at | Kimi K2.7 Code | Muse Spark Open (xhigh) |
| Open weights | Yes | Yes |
| Price in / out / cached input per 1M | $0.950 / $4.00 / $0.190 | $1.25 / $4.25 / $0.150 |
Category by category
Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.
| Category | Kimi K2.7 Code | Muse Spark Open (xhigh) | Leader |
|---|---|---|---|
| Reasoning | 53.0 | 51.3 | Kimi K2.7 Code |
| Coding | 51.9 | 36.0 | Kimi K2.7 Code |
| Agents & tools | 52.1 | – | not comparable |
| Maths | 51.9 | – | not comparable |
| Knowledge | 45.4 | 58.8 | Muse Spark Open |
| Instruction following | 62.9 | – | not comparable |
| Multimodal | – | 63.3 | not comparable |
| Long context | 64.3 | 63.8 | level |
On one chart
Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.
Not drawn, measured for fewer than two of these models: Agents & tools, Maths, Instruction following, Multimodal, Human preference.
Where sources disagree
Across the 19 benchmarks both have, Kimi K2.7 Code leads 4, Muse Spark Open leads 14 and 1 is a tie.
| Overall index | Kimi K2.7 Code | Muse Spark Open (xhigh) | Leader |
|---|---|---|---|
| BenchLeader Index | 53.9 | 53.5 | neck and neck |
Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).
Price, speed and latency
List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.
| Measure | Kimi K2.7 Code | Muse Spark Open (xhigh) |
|---|---|---|
| Blended price per 1M tokens | $1.71 | $2.00 |
| Input / output per 1M | $0.950 / $4.00 | $1.25 / $4.25 |
| Cached input per 1M | $0.190 | $0.150 |
| Price per task: chat reply | $0.0016 | $0.0018 |
| Price per task: summarize a 30-page report | $0.014 | $0.018 |
| Price per task: code edit | $0.012 | $0.014 |
| Price per task: agentic coding session | $0.073 | $0.092 |
| Price per task: structured extraction | $0.0027 | $0.0034 |
| Agentic coding session, input mostly cached | $0.039 | $0.043 |
| Cost per benchmark task (AA Index) | – | $2.02 |
| Output tokens per task (AA Index) | – | 66k |
| Of which reasoning | – | 32k |
| Time per task (AA Index) | – | 6 min |
| Output speed | 55 tok/s | 180 tok/s |
| Time to first answer | 29.6 s | 9.9 s |
| Time to a complete answer, last measured 13 Nov 2026 | 102.4 s | 39.9 s |
Each model is compared at its best-ranked reasoning-effort setting. Kimi K2.7 Code scores 53.9 and takes 30 s to its first answer token. Muse Spark Open (xhigh) scores 53.5 and takes 10 s to its first answer token. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.
Every benchmark
110 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.
| Metric | Publisher | Kimi K2.7 Code | Muse Spark Open (xhigh) |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 53.9 | 53.5 |
| Agents & tools score | BenchLeader | 52.1 | – |
| Coding score | BenchLeader | 51.9 | 36.0 |
| Instruction following score | BenchLeader | 62.9 | – |
| Knowledge score | BenchLeader | 45.4 | 58.8 |
| Long context score | BenchLeader | 64.3 | 63.8 |
| Maths score | BenchLeader | 51.9 | – |
| Reasoning score | BenchLeader | 53.0 | 51.3 |
| Multimodal score | BenchLeader | – | 63.3 |
| Blended price $/M | BenchLeader | $1.71 | $2.00 |
| Output speed | BenchLeader | 55 tok/s | 180 tok/s |
| Time to first answer | BenchLeader | 29.6 s | 9.9 s |
| Context window | BenchLeader | 262,144 tokens | 1,048,576 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 87.9% | – |
| SimpleBench | SimpleBench | 57.9% | – |
| LiveBench 2.0 Reasoning | LiveBench | – | 73.7%† |
| LiveBench 1.0 Reasoning | LiveBench | 82.8% | – |
| GPQA Diamond (AA) | Artificial Analysis | 89.6% | 89.0% |
| Humanity's Last Exam (AA) | Artificial Analysis | 35.0% | 43.9% |
| HLE-Diamond (AA) | Artificial Analysis | 14.8% | 26.3% |
| ARC-AGI-1 | ARC Prize | – | 89.0% |
| ARC-AGI-2 | ARC Prize | – | 57.5% |
| ARC-AGI-3 | ARC Prize | – | 0.4% |
| CritPt | Artificial Analysis | 10.0% | 14.9% |
| CritPt v2 hard subset (AA) | Artificial Analysis | – | 29.5% |
| HLE-Diamond (Scale) | Scale AI SEAL | – | 28.0%† |
| Chess Puzzles | Epoch AI Benchmarking Hub | 14.0% | – |
| Coding | |||
| SciCode | SciCode | 47.8% | – |
| WeirdML | WeirdML | 54.1% | – |
| FrontierCode | Cognition | 30.1% | – |
| Arena WebDev | Arena (LMArena) | 1474 | 1526† |
| LiveBench 2.0 Coding | LiveBench | – | 55.1%† |
| LiveBench 1.0 Coding | LiveBench | 74.0% | – |
| SciCode (AA) | Artificial Analysis | 47.8% | 58.5% |
| LiveCodeBench | Vals AI | 82.0% | – |
| SWE-bench (Vals) | Vals AI | 78.2% | – |
| Code Migration | Vals AI | 25.4% | 21.2% |
| Code Migration v1.1 | Vals AI | – | 26.9% |
Show all 110 benchmarks
| Metric | Publisher | Kimi K2.7 Code | Muse Spark Open (xhigh) |
|---|---|---|---|
| Coding | |||
| ProgramBench | Vals AI | 0.0% | – |
| Vibe Code Bench v1.1 | Vals AI | 47.2% | 76.2% |
| SWE Atlas v1.1 | Scale AI SEAL | – | 42.3%† |
| SWE Atlas v1.1: Codebase QnA | Scale AI SEAL | – | 50.0%† |
| SWE Atlas v1.1: Test Writing | Scale AI SEAL | – | 41.1%† |
| SWE Atlas v1.1: Refactoring | Scale AI SEAL | – | 35.7%† |
| SWE-Bench Pro V2 | Scale AI SEAL | – | 95.0% |
| SWE-Bench Pro V2 HARD | Scale AI SEAL | – | 82.3% |
| Surface Evolver Bench | Epoch AI Benchmarking Hub | 48.8% | – |
| DeepSWE v1.1 | Epoch AI Benchmarking Hub | 30.5% | – |
| ALE-Bench | Epoch AI Benchmarking Hub | 886.2 | – |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 1.0% | 8.1% |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | 67.4% | 76.4% |
| GameBench 2 Overall | DuelLab GameBench 2 | – | 55.3† |
| GameBench 2 by reasoning setting | DuelLab GameBench 2 | – | 59.3 |
| Design Arena Website | Design Arena | 1258 | 1305 |
| FrontierCode 1.1 (AA) | Artificial Analysis | – | 27.6% |
| FrontierCode 2.0 (AA) | Artificial Analysis | – | 7.6% |
| Agents & tools | |||
| APEX-Agents | Mercor | 37.6% | – |
| LiveBench Long Agentic | LiveBench | – | 12.1%† |
| LiveBench Browser | LiveBench | – | 26.2%† |
| LiveBench Agentic Coding | LiveBench | 45.7% | – |
| Terminal-Bench Hard | Artificial Analysis | 44.7% | – |
| τ²-Bench Telecom (AA) | Artificial Analysis | 90.1% | – |
| Terminal-Bench 2.1 (Vals) | Vals AI | 67.0% | – |
| HiL-Bench | Scale AI SEAL | – | 59.7%† |
| GDPval-AA v2.1, normalised score | Artificial Analysis | 26.9% | 45.1% |
| GDPval-AA v2.1 | Artificial Analysis | 1046 | 1401 |
| τ³-Banking (AA) | Artificial Analysis | 20.2% | – |
| Finance Agent v2 | Vals AI | – | 61.3% |
| Finance Agent v2.1 | Vals AI | – | 60.7% |
| Harvey's Legal Agent Benchmark v1.1 | Vals AI | – | 25.0% |
| Harvey's Legal Agent Benchmark v1 | Vals AI | – | 16.7% |
| Legal Research Bench | Vals AI | – | 38.5% |
| SkillsBench | Vals AI | 50.0% | – |
| Tax Agent Bench | Vals AI | – | 26.3% |
| Terminal-Bench 4.0 (Vals) | Vals AI | – | 6.6% |
| GBAEval | Epoch AI Benchmarking Hub | 0.8% | – |
| Vending-Bench 2 | Andon Labs | 5082.9 | – |
| AutomationBench | Artificial Analysis | 24.4% | 49.7% |
| GDP.pdf (AA) | Artificial Analysis | 11.2% | 18.6% |
| EnterpriseOps-Gym | Artificial Analysis | 40.2% | – |
| AA-Briefcase v1.1 | Artificial Analysis | 858 | 1280 |
| GDPx | GDPx by SemiAnalysis | – | 1.7%† |
| Design Arena Agentic | Design Arena | 1158 | – |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | – | 1.9% |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 54.0% | – |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 12.2% | – |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 95.6% | – |
| LiveBench Mathematics | LiveBench | 79.6% | – |
| ArXivMath | MathArena | – | 27.0% |
| BrokenArXiv | MathArena | – | 30.1% |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 36.5% | – |
| LiveBench 2.0 Data Analysis | LiveBench | – | 60.2%† |
| LiveBench Research | LiveBench | – | 82.7%† |
| LiveBench 1.0 Data Analysis | LiveBench | 62.7% | – |
| AA-Omniscience v1.1 | Artificial Analysis | – | 25.8 |
| PRBench Legal | Scale AI SEAL | – | 59.8%† |
| Excel Modeling Benchmark | Vals AI | – | 53.7% |
| Excel Modeling Benchmark v1.1 | Vals AI | – | 62.0% |
| DrugDiscoveryBench | Scale AI SEAL | 35.3%† | – |
| AA-Omniscience v1.1: accuracy | Artificial Analysis | – | 42.0% |
| AA-Omniscience v1.1: non-hallucination | Artificial Analysis | – | 72.0% |
| Instruction following | |||
| LiveBench Language | LiveBench | 77.9% | – |
| LiveBench Instruction Following | LiveBench | 56.3% | – |
| IFBench | Artificial Analysis | 63.1% | – |
| MultiChallenge | Scale AI SEAL | – | 74.2%† |
| Multimodal | |||
| LiveBench Design | LiveBench | – | 58.1%† |
| MMMU-Pro | Artificial Analysis | – | 80.9% |
| SAGE | Vals AI | – | 85.8% |
| Humanity's Sixth Sense | Scale AI SEAL | – | 29.6%† |
| Long context | |||
| AA-LCR v1.1 | Artificial Analysis | 79.3% | 78.3% |
| MLCR | Artificial Analysis | 12.2% | 7.8% |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 151.2 | 155.4† |
| LiveBench 2.0 | LiveBench | – | 48.5%† |
| LiveBench 1.0 | LiveBench | 68.4% | – |
| AA Intelligence Index v5.0.1 | Artificial Analysis | – | 24.9 |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 25.8 | 38.9 |
| AA Progress Index v1 | Artificial Analysis | 287.6 | 306.7 |
| Vals Index | Vals AI | – | 48.6 |
| Vals Finance Index | Vals AI | – | 53.0% |
| Vals Legal Index | Vals AI | – | 68.5% |
| Safety & honesty | |||
| FORTRESS | Scale AI SEAL | – | 7.6%† |
| DistressBench | Scale AI SEAL | – | 87.7%† |
| Metadata | |||
| AA Openness Index | Artificial Analysis | 27.8 | 27.8 |
Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.
DuelLab GameBench 2, release gb2-20261111-7b598bb13de4 (https://benchmarks.duellab.org/releases/gb2-20261111-7b598bb13de4/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.
What changed
- Kimi K2.7 Code: 1 score revision.
- Muse Spark Open (xhigh): 1 score revision.
- Kimi K2.7 Code: 1 score revision.
- Muse Spark Open (xhigh): 1 new result (GameBench 2 by reasoning setting); 1 score revision.
- Muse Spark Open (xhigh): BenchLeader Index 53.3 → 53.8.
- Kimi K2.7 Code: 1 new result (AA-Briefcase v1.1).
- Muse Spark Open (xhigh): 1 new result (ARC-AGI-3).
- Kimi K2.7 Code: BenchLeader Index 55.9 → 54.2.
- Muse Spark Open (xhigh): BenchLeader Index 59.5 → 53.3.
- Kimi K2.7 Code: 2 new results (AA Progress Index v1, HLE-Diamond (AA)).
Related comparisons
- Kimi K2.7 Code vs MiniMax M3.1 Flash
- Muse Spark Open vs Claude Opus 5.5
- Muse Spark Open vs GPT-6 Astra
- Muse Spark Open vs GPT-6.1 Sol
- Muse Spark Open vs Claude Fable 5
- Muse Spark Open vs Gemini 4 Argon
- Muse Spark Open vs GPT-5.6 Sol
- Muse Spark Open vs Claude Sonnet 5.5
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
Kimi K2.7 Code vs Muse Spark Open: questions
- Is Kimi K2.7 Code better than Muse Spark Open?
- Kimi K2.7 Code and Muse Spark Open (xhigh) are neck and neck on quality (53.9 vs 53.5, within the ±8.0 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 16 Nov 2026, so check the category scores for your use.
- Kimi K2.7 Code vs Muse Spark Open: which is cheaper, and will that change?
- Kimi K2.7 Code is cheaper today: $1.71 against $2.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Is Kimi K2.7 Code or Muse Spark Open better for coding?
- Kimi K2.7 Code scores higher in coding (52 vs 36 on the category index, where 50 is average).
- Which is faster, Kimi K2.7 Code or Muse Spark Open?
- Muse Spark Open streams 3×+ faster (180 vs 55 output tokens per second, Artificial Analysis 7-day median).
- Which is better for long documents?
- They are level on long-context benchmarks (64 vs 64). Muse Spark Open accepts more context: 1,048,576 against 262,144 tokens for Kimi K2.7 Code.
More questions (3)
- Which has lower latency?
- Muse Spark Open answers first: 9.9 s against 29.6 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
- Do independent benchmarks agree on Kimi K2.7 Code vs Muse Spark Open?
- Across the 19 benchmarks both have, Kimi K2.7 Code leads 4, Muse Spark Open leads 14 and 1 is a tie.
- When was Muse Spark Open released?
- Muse Spark Open was released on 15 Oct 2026.