GLM 5 Turbo vs Muse Spark
Full profiles: GLM 5 Turbo (Zhipu AI) and Muse Spark (Meta): every benchmark, price and rank.
Updated · Data from 8 sources
As of 16 Nov 2026, there are not enough shared benchmarks for a verdict. Muse Spark leads in every category we measure both in.
Key facts (as of 16 Nov 2026)
- GLM 5 Turbo costs $1.90 per 1M tokens blended (list price, 16 Nov 2026).
- Of the 9 benchmarks both models have, Muse Spark leads 7 (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 |
|---|---|---|---|
| Agents & tool use | Muse Spark | 65 vs 63 category score; Terminal-Bench Hard 45.5% vs 33.3% | Medium |
| Long documents | Muse Spark | 64 vs 60 category score; AA-LCR v1.1 78.0% vs 71.7% | Medium |
| Following instructions | Muse Spark | 77 vs 72 category score; IFBench 75.9% vs 73.2% | Medium |
| Chat quality (human preference) | – | Only Muse Spark has an Arena rating (68 category score) | Not comparable |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- Not enough shared benchmarks for a verdict.
- Muse Spark is stronger in agents & tools, instruction following, long context.
At a glance
| Spec | GLM 5 Turbo | Muse Spark |
|---|---|---|
| Developer | Zhipu AI | Meta |
| Released | 15 Mar 2026 | 8 Apr 2026 |
| Context window (tokens) | 200,000 | 262,144 |
| Compared at | GLM 5 Turbo | Muse Spark |
| Open weights | No | No |
| Price in / out / cached input per 1M | $1.20 / $4.00 / $0.240 | – |
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 | GLM 5 Turbo | Muse Spark | Leader |
|---|---|---|---|
| Reasoning | – | 67.4 | not comparable |
| Coding | – | 65.1 | not comparable |
| Agents & tools | 62.8 | 65.0 | Muse Spark |
| Maths | – | 59.1 | not comparable |
| Knowledge | – | 63.1 | not comparable |
| Instruction following | 71.7 | 76.9 | Muse Spark |
| Human preference | – | 68.1 | not comparable |
| Multimodal | – | 64.2 | not comparable |
| Long context | 60.3 | 63.6 | Muse Spark |
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: Reasoning, Coding, Maths, Knowledge, Multimodal, Human preference.
Where sources disagree
Across the 9 benchmarks both have, GLM 5 Turbo leads 2, Muse Spark leads 7.
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 | GLM 5 Turbo | Muse Spark |
|---|---|---|
| Blended price per 1M tokens | $1.90 | – |
| Input / output per 1M | $1.20 / $4.00 | – / – |
| Cached input per 1M | $0.240 | – |
| Price per task: chat reply | $0.0017 | – |
| Price per task: summarize a 30-page report | $0.017 | – |
| Price per task: code edit | $0.013 | – |
| Price per task: agentic coding session | $0.088 | – |
| Price per task: structured extraction | $0.0032 | – |
| Agentic coding session, input mostly cached | $0.045 | – |
| Output speed | 26 tok/s | – |
| Time to first answer | 4.2 s | – |
Each model is compared at its best-ranked reasoning-effort setting. GLM 5 Turbo and takes 4 s to its first answer token. Muse Spark scores 64.2. 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
55 benchmarks, grouped by category; each name links to the board and its leaders.
| Metric | Publisher | GLM 5 Turbo | Muse Spark |
|---|---|---|---|
| BenchLeader Index | BenchLeader | – | 64.2 |
| Agents & tools score | BenchLeader | 62.8 | 65.0 |
| Instruction following score | BenchLeader | 71.7 | 76.9 |
| Long context score | BenchLeader | 60.3 | 63.6 |
| Coding score | BenchLeader | – | 65.1 |
| Human preference score | BenchLeader | – | 68.1 |
| Knowledge score | BenchLeader | – | 63.1 |
| Maths score | BenchLeader | – | 59.1 |
| Multimodal score | BenchLeader | – | 64.2 |
| Reasoning score | BenchLeader | – | 67.4 |
| Blended price $/M | BenchLeader | $1.90 | – |
| Output speed | BenchLeader | 26 tok/s | – |
| Time to first answer | BenchLeader | 4.2 s | – |
| Context window | BenchLeader | 200,000 tokens | 262,144 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | – | 89.8% |
| Humanity's Last Exam | Scale AI / CAIS | – | 40.6% |
| Arena Hard Prompts | Arena (LMArena) | – | 1506 |
| GPQA Diamond (AA) | Artificial Analysis | 84.8% | 88.4% |
| Humanity's Last Exam (AA) | Artificial Analysis | 27.8% | 40.7% |
| GPQA Diamond (Vals) | Vals AI | – | 89.7% |
| CritPt | Artificial Analysis | 0.3% | 11.3% |
| Arena Expert | Arena (LMArena) | – | 1491 |
| Coding | |||
| SciCode | SciCode | – | 51.5% |
| Arena Coding | Arena (LMArena) | – | 1529 |
| SWE-bench (Vals) | Vals AI | – | 74.4% |
| SWE-Bench Pro | Scale AI SEAL | – | 55.0% |
| Vibe Code Bench v1.1 | Vals AI | – | 19.7% |
| SWE-Bench Pro (private) | Scale AI SEAL | – | 44.7% |
| SWE Atlas v1.0: Codebase QnA | Scale AI SEAL | – | 24.2% |
| SWE Atlas v1.0: Test Writing | Scale AI SEAL | – | 31.1% |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | – | 62.2% |
| Design Arena Website | Design Arena | 1254 | 1253 |
| Agents & tools | |||
| Terminal-Bench Hard | Artificial Analysis | 33.3% | 45.5% |
| τ²-Bench Telecom (AA) | Artificial Analysis | 98.5% | 91.5% |
| MCP Atlas | Scale AI SEAL | – | 82.2% |
| GDPval-AA v2.1 | Artificial Analysis | – | 1006 |
| Terminal-Bench 2.0 (Vals) | Vals AI | – | 59.5% |
| AA-Briefcase v1.1 | Artificial Analysis | – | 646 |
| Maths | |||
| OTIS Mock AIME | Epoch AI Benchmarking Hub | – | 88.9% |
Show all 55 benchmarks
| Metric | Publisher | GLM 5 Turbo | Muse Spark |
|---|---|---|---|
| Maths | |||
| ProofBench | Vals AI | – | 17.0% |
| AIME (Vals) | Vals AI | – | 96.9% |
| Arena Maths | Arena (LMArena) | – | 1469 |
| Knowledge | |||
| MMLU-Pro | Vals AI | – | 87.3% |
| LegalBench | Vals AI | – | 84.2% |
| CorpFin | Vals AI | – | 65.1% |
| TaxEval | Vals AI | – | 77.7% |
| PRBench Finance | Scale AI SEAL | – | 52.4% |
| PRBench Legal | Scale AI SEAL | – | 52.3% |
| MultiNRC | Scale AI SEAL | – | 59.0% |
| CaseLaw v2 | Vals AI | – | 63.1% |
| MedCode | Vals AI | – | 51.3% |
| MedScribe | Vals AI | – | 85.9% |
| Instruction following | |||
| IFBench | Artificial Analysis | 73.2% | 75.9% |
| MultiChallenge | Scale AI SEAL | – | 75.5% |
| TutorBench | Scale AI SEAL | – | 68.5% |
| Arena Instruction Following | Arena (LMArena) | – | 1465 |
| Human preference | |||
| Arena Text | Arena (LMArena) | – | 1489 |
| Arena Creative Writing | Arena (LMArena) | – | 1464 |
| Arena Multi-turn | Arena (LMArena) | – | 1493 |
| Arena Longer Queries | Arena (LMArena) | – | 1476 |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | – | 1306 |
| MMMU-Pro | Artificial Analysis | – | 80.5% |
| MMMU-Pro (Vals) | Vals AI | – | 87.4% |
| Long context | |||
| AA-LCR v1.1 | Artificial Analysis | 71.7% | 78.0% |
| Arena Document | Arena (LMArena) | – | 1467 |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | – | 153.1 |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 26.6 | 31.3 |
| AA Progress Index v1 | Artificial Analysis | 277.3 | 289.1 |
| Safety & honesty | |||
| FORTRESS | Scale AI SEAL | – | 20.2% |
Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.
What changed
- Muse Spark: 2 new results (GDPval-AA v2.1, AA-Briefcase v1.1).
- Muse Spark: BenchLeader Index 64.6 → 64.1.
- Muse Spark: 1 score revision.
- GLM 5 Turbo: 1 new result (AA Progress Index v1).
- Muse Spark: 1 new result (AA Progress Index v1).
Related comparisons
- Muse Spark vs Claude Fable 5.2
- Muse Spark vs Gemini 4 Carbon
- Muse Spark vs Claude Opus 5.5
- Muse Spark vs GPT-6 Astra
- Muse Spark vs Claude Fable 5.1
- Muse Spark vs GPT-6.1 Sol
- Muse Spark vs Claude Fable 5
- Muse Spark vs GPT-5.6 Sol
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
GLM 5 Turbo vs Muse Spark: questions
- Is GLM 5 Turbo or Muse Spark better for agents and tool use?
- Muse Spark scores higher in agentic tasks (65 vs 63 on the category index, where 50 is average). On Terminal-Bench Hard, the top shared agents & tools benchmark, it is 33.3% for GLM 5 Turbo against 45.5% for Muse Spark.
- Which is better for long documents?
- Muse Spark scores higher on long-context benchmarks (64 vs 60). Muse Spark accepts more context: 262,144 against 200,000 tokens for GLM 5 Turbo.
- Do independent benchmarks agree on GLM 5 Turbo vs Muse Spark?
- Across the 9 benchmarks both have, GLM 5 Turbo leads 2, Muse Spark leads 7.
- When was Muse Spark released?
- Muse Spark was released on 8 Apr 2026.