vs
1:8
overall capability wins (1 ties across 10 scored). Muse Spark 1.3 leads overall.
- Reasoning1:4
- Coding0:2
- Agents0:1
- Tool use0:1
- Composite indices0:2
Verdict
Muse Spark 1.3 leads coding (widest gap: +5.7 on TermBench).
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Muse Spark 1.2
- $2.31
- Muse Spark 1.3
- $2.31
Price per 1M tokens
Full comparison
Organization
Meta
Meta
Family
Llama / Muse
Llama / Muse
License
Proprietary
Proprietary
Open weights
No
No
Release
Aug 5, 2026
Sep 2, 2026
Knowledge cutoff
-
-
API / provider
Meta
Meta
Modalities
text, image, video, audio, pdf → text
text, image, video, pdf → text
Specs
Context window
1.0M
1.0M
Max output
131K
-
Parameters
—
—
Pricing
Input $/1M
$1.25
$1.25
Output $/1M
$4.25
$4.25
Blended $/1M (3∶1)
$2
$2
Speed
tok/s
95
-
TTFT (s)
0.38
-
Reasoning
MMLU-Pro
—
—
GPQA Diamond
90.4
93.5
Humanity's Last Exam
45.5
48.7
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
45.4
43.6
AA-LCR v1.1
79.0
83.0
CritPt
17.7
24.9
MMMU-Pro
—
—
IFBench
—
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
80.1
85.8
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
57.4
58.8
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
—
—
DeepSWE
55.0
—
WebDev Arena
1,535
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,500
—
| Metric | Muse Spark 1.2 | Muse Spark 1.3 | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Meta | Meta | — |
| Family | Llama / Muse | Llama / Muse | — |
| License | Proprietary | Proprietary | — |
| Open weights | No | No | — |
| Release | Aug 5, 2026 | Sep 2, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | Meta | Meta | — |
| Modalities | text, image, video, audio, pdf → text | text, image, video, pdf → text | — |
| Specs | |||
| Context window | 1.0M | 1.0M | tie |
| Max output | 131K | - | — |
| Parameters | — | — | — |
| Pricing | |||
| Input $/1M | $1.25 | $1.25 | tie |
| Output $/1M | $4.25 | $4.25 | tie |
| Blended $/1M (3∶1) | $2 | $2 | tie |
| Speed | |||
| tok/s | 95 | - | — |
| TTFT (s) | 0.38 | - | — |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | 90.4 | 93.5 | B +3.1 pts |
| Humanity's Last Exam | 45.5 | 48.7 | B +3.2 pts |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 45.4 | 43.6 | A +1.8 pts |
| AA-LCR v1.1 | 79.0 | 83.0 | B +4.0 pts |
| CritPt | 17.7 | 24.9 | B +7.2 pts |
| MMMU-Pro | — | — | — |
| IFBench | — | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | 80.1 | 85.8 | B +5.7 pts |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | 57.4 | 58.8 | B +1.4 pts |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | 55.0 | — | — |
| WebDev Arena | 1,535 | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,500 | — | — |
Muse Spark 1.2
- Context
- 1.0M / 131K out
- Parameters
- —
- Price
- $1.25 / $4.25
- Speed
- 95 tok/s · 0.38s TTFT
- Modalities
- text, image, video, audio, pdf → text
- License
- Proprietary
Co-trained with Muse Code for higher first-attempt accuracy and more reliable tool calling — 1M context coding-focused upgrade; Meta Model API at $1.25/$4.25 (contributor tier $0.10/$0.20).
Muse Spark 1.3
- Context
- 1.0M
- Parameters
- —
- Price
- $1.25 / $4.25
- Speed
- —
- Modalities
- text, image, video, pdf → text
- License
- Proprietary
Meta's Sept 2026 coding/agentic upgrade to Muse Spark — ~20% fewer tool calls and ~25% fewer tokens than 1.2 in internal comparisons, with longer-horizon multi-workflow persistence and more proactive clarifying questions. Same 1M context and $1.25/$4.25 Meta Model API pricing as 1.2 (contributor tier $0.10/$0.20). Open weights planned but not yet released; max reasoning mode pending further safety testing.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.