SShogo Bench

GPT-6 Astra vs Muse Spark 1.3

GPT-6 Astra leads Muse Spark 1.3 by 16.2 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. GPT-6 Astra wins 14 and Muse Spark 1.3 wins 1 of 15 shared benchmarks. Muse Spark 1.3 is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarGPT-6 AstraMuse Spark 1.3
Agentic coding8060
Long-horizon autonomy8064
Knowledge-work agents5645
Computer use6743
Security & context learning6353

Shared benchmarks

BenchmarkGPT-6 AstraMuse Spark 1.3
APEX-Agentsindex64.7%58.6%
FrontierSWEindex65.5%25.8%
Chess Puzzles76.0%37.0%
CritPt31.7%26.0%
DTBench97.3%96.5%
FrontierMath Tier 497.6%46.3%
FrontierMath Tiers 1-393.7%74.4%
LMCA64.4%53.9%
MirrorCode60.0%54.4%
OTIS Mock AIME100.0%99.2%
ProofBench99.0%58.0%
SciCode56.5%59.7%
SimpleBench86.5%81.8%
WebDev Arena18001652
WeirdML v342.2%6.7%

Specs and price

GPT-6 AstraMuse Spark 1.3
Agent Index70.9 (#4)54.7 (#28)
Agent $/M$4.62$0.030
Input / output $/M$10.00 / $50.00$0.10 / $0.20
Context1.1M1M
ReleasedSep 4, 2026Sep 2, 2026
WeightsClosedClosed
API hosts62

Don’t pick one. Route.

Shogo can run GPT-6 Astra and Muse Spark 1.3 side by side and send each step of an agent to the one that’s cheaper for it.

Try both in Shogo