In short
How they differ.
Worked out from the figures below. Prices and specs are the makers' own, read Oct 20, 2026.
- Muse Spark 1.3 reads more at once: a context window of 1M tokens against 256K.
- Gemma 4 31B can be downloaded and run on your own machine; Muse Spark 1.3 can't.
- No independent board we follow ranks both models yet.
Side by side
The specs.
| Gemma 4 31B | Muse Spark 1.3 | |
|---|---|---|
| Made by | Google · US | Meta · US |
| Released | Apr 2, 2026 | Not stated |
| Price per 1M tokens | Free to download; hosts set prices | $1.25 in / $4.25 out |
| Context window | 256K | 1M |
| Takes in | text, image, video | text, image, video, audio, pdf |
| Download and run it yourself | Yes (Apache 2.0) | No |
| Size | 30.7B (dense) parameters | Not published |
| Maker says | “Today, we are introducing Gemma 4 — our most intelligent open models to date.” | “Muse Spark 1.3 is trained for agentic workflows and optimized for competitive coding performance.” |
| Sources | Prices, Oct 20, 2026 · Description | Prices, Oct 20, 2026 · Description |
Independent tests
Where they place.
Position on each board we follow (1 is best). Epoch AI's and Vectara's scores are shown; the other boards are linked for their own figures.
| Board | Gemma 4 31B | Muse Spark 1.3 |
|---|---|---|
| Epoch Capabilities IndexEpoch AI, Oct 20, 2026 | Not listed | #22156.76 |
| LMArena textLMArena, Oct 15, 2026 | Not listed | #10 |
| LMArena coding chatsLMArena, Oct 15, 2026 | Not listed | #15 |
| LMArena WebDevLMArena, Oct 19, 2026 | Not listed | #14 |
| LMArena mathLMArena, Oct 15, 2026 | Not listed | #11 |
| LMArena visionLMArena, Oct 19, 2026 | Not listed | #13 |
| LMArena documentsLMArena, Sep 13, 2026 | Not listed | #15 |
| LMArena Agent ArenaLMArena, Oct 19, 2026 | Not listed | #16 |
| Humanity's Last ExamScale AI | Not listed | #12 |
| Vibe Code BenchVals AI, Oct 18, 2026 | Not listed | #14 |
| CUA-bench (games)Vals AI, Oct 13, 2026 | Not listed | #7 |
More comparisons