Muse SparkvsGemini 3.1 Pro Preview
Across 8 shared benchmarks, Gemini 3.1 Pro Preview leads overall: Muse Spark wins 2, Gemini 3.1 Pro Preview wins 6, with 0 ties and an average score difference of -6.60.
Facebook AI研究实验室 · 2026-04-08 · Reasoning model
Google Deep Mind · 2026-02-20 · Multimodal model
Benchmark scores
Grouped by capability, sorted by largest gap within each. 8 shared benchmarks.
General Knowledge
Gemini 3.1 Pro Preview 2/3| Benchmark | Muse Spark | Gemini 3.1 Pro Preview | Diff |
|---|---|---|---|
| ARC-AGI-2 | 42.5025 / 59Thinking (No Tools) | 77.107 / 59Thinking High (No Tools) | -34.60 |
| HLE | 584 / 157深度思考（无工具、并行） | 51.4015 / 157Thinking High (With Tools) | +6.60 |
| GPQA Diamond | 89.5022 / 178Thinking (No Tools) | 94.303 / 178Thinking High (No Tools) | -4.80 |
Math and Reasoning
Even 2/2| Benchmark | Muse Spark | Gemini 3.1 Pro Preview | Diff |
|---|---|---|---|
| FrontierMath - Tier 4 | 14.6023 / 80Normal (No Tools) | 16.7020 / 80Normal (No Tools) | -2.10 |
| FrontierMath | 399 / 60Thinking (No Tools) | 36.9011 / 60Thinking High (No Tools) | +2.10 |
Agent Level Benchmark
Gemini 3.1 Pro Preview 1/1| Benchmark | Muse Spark | Gemini 3.1 Pro Preview | Diff |
|---|---|---|---|
| τ²-Bench - Telecom | 9220 / 35Thinking (With Tools) | 99.301 / 35Thinking High (With Tools) | -7.30 |
AI Agent - Tool Usage
Gemini 3.1 Pro Preview 1/1| Benchmark | Muse Spark | Gemini 3.1 Pro Preview | Diff |
|---|---|---|---|
| Terminal Bench 2.0 | 5924 / 46Thinking (With Tools) | 68.508 / 46Thinking High (With Tools) | -9.50 |
Coding and Software Engineer
Gemini 3.1 Pro Preview 1/1| Benchmark | Muse Spark | Gemini 3.1 Pro Preview | Diff |
|---|---|---|---|
| SWE-bench Verified | 77.4024 / 108Thinking (With Tools) | 80.6010 / 108Thinking High (With Tools) | -3.20 |
Specs
| Field | Muse Spark | Gemini 3.1 Pro Preview |
|---|---|---|
| Publisher | Facebook AI研究实验室 | Google Deep Mind |
| Release date | 2026-04-08 | 2026-02-20 |
| Model type | Reasoning model | Multimodal model |
| Architecture | Dense | Dense |
| Parameters | Not available | Not available |
| Context length | 262K | 1M |
| Max output | Not available | 32K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Muse Spark | Gemini 3.1 Pro Preview |
|---|---|---|
| Text input | Not public | $2 / 1M tokens |
| Text output | Not public | $12 / 1M tokens |
One or both models have incomplete public pricing.
Summary
- Gemini 3.1 Pro Previewleads in:General Knowledge (2/3), Agent Level Benchmark (1/1), AI Agent - Tool Usage (1/1), Coding and Software Engineer (1/1)
- Tied in:Math and Reasoning
On average across the 8 shared benchmarks, Gemini 3.1 Pro Preview scores 6.60 higher.
Largest single-benchmark gap: ARC-AGI-2 — Muse Spark 42.50 vs Gemini 3.1 Pro Preview 77.10 (-34.60).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.