Compare / head-to-head
Gemini 3 Pro Previewvs
Muse Spark 1.3
Muse Spark 1.3 leads 3 of 3 shared benchmarks. Muse Spark 1.3 is 2.5x cheaper per token. Both offer a 1.0M-token context window.
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
0 – 3
Muse Spark 1.3 leads
Cheaper per token
Muse Spark 1.3
2.5x cheaper, input + output
Larger context
Tie
both 1.0M tokens
Providers
2 providers
Google · Meta
Gemini 3 Pro Preview
Google · released 2025-11-18
textvisionaudiovideo
Muse Spark 1.3
Meta · released 2026-09-02
textvisionvideo
Quality
Benchmark matrix
3 shared · 21 only Gemini 3 Pro Preview · 26 only Muse Spark 1.3| Benchmark | Gemini 3 Pro Preview | Muse Spark 1.3 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 3 | ||||
| Chess Puzzles (Epoch AI run) | 31% ↗epoch run | 38% ↗epoch run | -7 pt | Muse Spark 1.3 |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 91.4% ↗epoch run | 99.2% ↗epoch run | -7.8 pt | Muse Spark 1.3 |
| SimpleBench (SimpleBench) | 76.4% ↗ | 81.8% ↗ | -5.4 pt | Muse Spark 1.3 |
| Only Gemini 3 Pro Preview reports · 21 | ||||
| SWE-bench Verified | 76.2% ↗ | not reported | — | — |
| GPQA Diamond | 91.9% ↗ | not reported | — | — |
| AIME 2025 | 95% ↗ | not reported | — | — |
| Humanity's Last Exam (no tools) | 37.5% ↗ | not reported | — | — |
| AIME 2025 (with code execution) | 100% ↗ | not reported | — | — |
| ARC-AGI-2 (ARC Prize Verified) | 31.1% ↗ | not reported | — | — |
| BALROG (BALROG) | 58.1% ↗ | not reported | — | — |
| CharXiv Reasoning | 81.4% ↗ | not reported | — | — |
| GSO Opt@1 (GSO) | 17.7% ↗ | not reported | — | — |
| Humanity's Last Exam (with search and code execution) | 45.8% ↗ | not reported | — | — |
| MathArena Apex | 23.4% ↗ | not reported | — | — |
| MMMLU | 91.8% ↗ | not reported | — | — |
| MMMU-Pro | 81% ↗ | not reported | — | — |
| SAGE (Vals AI) | 47.6% ↗ | not reported | — | — |
| SimpleQA Verified | 72.1% ↗ | not reported | — | — |
| SWE-bench Verified (Epoch AI run) | 72.9% ↗epoch run | not reported | — | — |
| Terminal-Bench 2.0 (Terminus-2 agent) | 54.2% ↗ | not reported | — | — |
| Vending-Bench 2 (Andon Labs) | 5478.16 ↗ | not reported | — | — |
| Video-MMMU | 87.6% ↗ | not reported | — | — |
| WeirdML (Håvard Tveit Ihle) | 69.9% ↗ | not reported | — | — |
| τ2-bench | 85.4% ↗ | not reported | — | — |
| Only Muse Spark 1.3 reports · 26 | ||||
| LMArena Elo | not reported | 1494.3 ↗ | — | — |
| APEX-Agents (Mercor) | not reported | 57.8% ↗ | — | — |
| AutomationBench | not reported | 49.6% ↗ | — | — |
| DeepSearchQA | not reported | 90.3% ↗ | — | — |
| DeepSWE v1.1 | not reported | 75.4% ↗ | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | not reported | 46.3% ↗epoch run | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | not reported | 74.4% ↗epoch run | — | — |
| GDPval-AA v2 | not reported | 1754 ↗ | — | — |
| JobBench | not reported | 64.9% ↗ | — | — |
| LiveBench Agentic Coding (LiveBench) | not reported | 64.1% ↗ | — | — |
| LiveBench Coding (LiveBench) | not reported | 81.1% ↗ | — | — |
| LiveBench Data Analysis (LiveBench) | not reported | 79.6% ↗ | — | — |
| LiveBench Instruction Following (LiveBench) | not reported | 78% ↗ | — | — |
| LiveBench Language (LiveBench) | not reported | 82.8% ↗ | — | — |
| LiveBench Mathematics (LiveBench) | not reported | 95.9% ↗ | — | — |
| LiveBench Reasoning (LiveBench) | not reported | 89.7% ↗ | — | — |
| LMArena Agent (LMArena) | not reported | 0.04 ↗ | — | — |
| LMArena Vision (LMArena) | not reported | 1290.2 ↗ | — | — |
| LMArena WebDev (LMArena) | not reported | 1656.6 ↗ | — | — |
| MRCR 256K-512K | not reported | 98.5% ↗ | — | — |
| MRCR 512K-1M | not reported | 98.1% ↗ | — | — |
| Mystery Game Puzzles (Epoch AI run) | not reported | 25% ↗epoch run | — | — |
| OSWorld 2.0 (partial) | not reported | 66.9% ↗ | — | — |
| SWE-Atlas Codebase QnA | not reported | 59.4% ↗ | — | — |
| Terminal-Bench 2.1 | not reported | 88.8% ↗ | — | — |
| Terminal-Bench 4.0 (Vals AI) | not reported | 24.7% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Gemini 3 Pro Preview minus Muse Spark 1.3 in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Gemini 3 Pro Preview | Muse Spark 1.3 | Edge |
|---|---|---|---|
| Context window | 1.0M tokens | 1.0M tokens | Tie |
| Max output | 66K tokens | — | — |
| Input price / 1M | $2 ↗ | $1.25 ↗ | Muse Spark 1.3 |
| Output price / 1M | $12 ↗ | $4.25 ↗ | Muse Spark 1.3 |
| Cached input / 1M | — | $0.15 ↗ | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $14 | $5.5 | Muse Spark 1.3 |
| Modalities | text · vision · audio · video | text · vision · video | Gemini 3 Pro Preview |
| Released | 2025-11-18 | 2026-09-02 | — |
| Cited benchmark scores | 25 | 29 | — |
Reliability
Provider status
Live from /statusMore matchups
Gemini 3 Pro Preview vs …
Models sharing the most benchmarksMore matchups
Muse Spark 1.3 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Gemini 3 Pro Preview and Muse Spark 1.3 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.