Compare / head-to-head
Muse Spark 1.3vs
Qwen3 235B-A22B Thinking 2507
Muse Spark 1.3 leads 4 of 4 shared benchmarks. Qwen3 235B-A22B Thinking 2507 is 2.2x cheaper per token. Muse Spark 1.3 has the larger context window (1.0M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
4 – 0
Muse Spark 1.3 leads
Cheaper per token
Qwen3 235B-A22B Thinking 2507
2.2x cheaper, input + output
Larger context
Muse Spark 1.3
1.0M tokens
Providers
2 providers
Meta · Qwen
Muse Spark 1.3
Meta · released 2026-09-02
textvisionvideo
Qwen3 235B-A22B Thinking 2507
Qwen · released 2025-07-25
text
Quality
Benchmark matrix
4 shared · 25 only Muse Spark 1.3 · 19 only Qwen3 235B-A22B Thinking 2507| Benchmark | Muse Spark 1.3 | Qwen3 235B-A22B Thinking 2507 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 4 | ||||
| LMArena Elo | 1494.3 ↗ | 1400 ↗ | +94.3 | Muse Spark 1.3 |
| Chess Puzzles (Epoch AI run) | 38% ↗epoch run | 12% ↗epoch run | +26 pt | Muse Spark 1.3 |
| Mystery Game Puzzles (Epoch AI run) | 25% ↗epoch run | 9% ↗epoch run | +16 pt | Muse Spark 1.3 |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 99.2% ↗epoch run | 86.7% ↗epoch run | +12.5 pt | Muse Spark 1.3 |
| Only Muse Spark 1.3 reports · 25 | ||||
| APEX-Agents (Mercor) | 57.8% ↗ | not reported | — | — |
| AutomationBench | 49.6% ↗ | not reported | — | — |
| DeepSearchQA | 90.3% ↗ | not reported | — | — |
| DeepSWE v1.1 | 75.4% ↗ | not reported | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | 46.3% ↗epoch run | not reported | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 74.4% ↗epoch run | not reported | — | — |
| GDPval-AA v2 | 1754 ↗ | not reported | — | — |
| JobBench | 64.9% ↗ | not reported | — | — |
| LiveBench Agentic Coding (LiveBench) | 64.1% ↗ | not reported | — | — |
| LiveBench Coding (LiveBench) | 81.1% ↗ | not reported | — | — |
| LiveBench Data Analysis (LiveBench) | 79.6% ↗ | not reported | — | — |
| LiveBench Instruction Following (LiveBench) | 78% ↗ | not reported | — | — |
| LiveBench Language (LiveBench) | 82.8% ↗ | not reported | — | — |
| LiveBench Mathematics (LiveBench) | 95.9% ↗ | not reported | — | — |
| LiveBench Reasoning (LiveBench) | 89.7% ↗ | not reported | — | — |
| LMArena Agent (LMArena) | 0.04 ↗ | not reported | — | — |
| LMArena Vision (LMArena) | 1290.2 ↗ | not reported | — | — |
| LMArena WebDev (LMArena) | 1656.6 ↗ | not reported | — | — |
| MRCR 256K-512K | 98.5% ↗ | not reported | — | — |
| MRCR 512K-1M | 98.1% ↗ | not reported | — | — |
| OSWorld 2.0 (partial) | 66.9% ↗ | not reported | — | — |
| SimpleBench (SimpleBench) | 81.8% ↗ | not reported | — | — |
| SWE-Atlas Codebase QnA | 59.4% ↗ | not reported | — | — |
| Terminal-Bench 2.1 | 88.8% ↗ | not reported | — | — |
| Terminal-Bench 4.0 (Vals AI) | 24.7% ↗ | not reported | — | — |
| Only Qwen3 235B-A22B Thinking 2507 reports · 19 | ||||
| GPQA Diamond | not reported | 80.1% ↗epoch run | — | — |
| MMLU-Pro | not reported | 84.4% ↗ | — | — |
| AIME25 | not reported | 92.3% ↗ | — | — |
| Arena-Hard v2 | not reported | 79.7% ↗ | — | — |
| BFCL-v3 | not reported | 71.9% ↗ | — | — |
| GPQA | not reported | 81.1% ↗ | — | — |
| HLE (text-only) | not reported | 18.2% ↗ | — | — |
| HMMT25 | not reported | 83.9% ↗ | — | — |
| IFEval | not reported | 87.8% ↗ | — | — |
| LiveBench 20241125 | not reported | 78.4% ↗ | — | — |
| LiveCodeBench v6 (25.02-25.05) | not reported | 74.1% ↗ | — | — |
| MMLU-Redux | not reported | 93.8% ↗ | — | — |
| OJBench | not reported | 32.5% ↗ | — | — |
| SimpleQA Verified | not reported | 40.4% ↗epoch run | — | — |
| SuperGPQA | not reported | 64.9% ↗ | — | — |
| TAU2-Airline | not reported | 58% ↗ | — | — |
| TAU2-Retail | not reported | 71.9% ↗ | — | — |
| Vending-Bench 2 (Andon Labs) | not reported | -11.34 ↗ | — | — |
| WeirdML (Håvard Tveit Ihle) | not reported | 41% ↗ | — | — |
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 Muse Spark 1.3 minus Qwen3 235B-A22B Thinking 2507 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 | Muse Spark 1.3 | Qwen3 235B-A22B Thinking 2507 | Edge |
|---|---|---|---|
| Context window | 1.0M tokens | 262K tokens | Muse Spark 1.3 |
| Max output | — | — | — |
| Input price / 1M | $1.25 ↗ | $0.23 ↗ | Qwen3 235B-A22B Thinking 2507 |
| Output price / 1M | $4.25 ↗ | $2.3 ↗ | Qwen3 235B-A22B Thinking 2507 |
| Cached input / 1M | $0.15 ↗ | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $5.5 | $2.53 | Qwen3 235B-A22B Thinking 2507 |
| Modalities | text · vision · video | text | Muse Spark 1.3 |
| Released | 2026-09-02 | 2025-07-25 | — |
| Cited benchmark scores | 29 | 23 | — |
Reliability
Provider status
Live from /statusMore matchups
Muse Spark 1.3 vs …
Models sharing the most benchmarksMore matchups
Qwen3 235B-A22B Thinking 2507 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Muse Spark 1.3 and Qwen3 235B-A22B Thinking 2507 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.