Measured · release 2026-10-12
Muse Spark code review, measured
Muse Spark 1.3 reviewed the same pull requests as every other model on 8 of IonWarp’s 8 benchmark lanes, each finding matched against the pull request’s known issues. Muse Spark 1.3 is the best model we ran on Code Review (F1 56.6%) and API Performance (F1 56.0%). On general code review it costs $0.424 a review; DeepSeek V4.1 Flash, which IonWarp’s Code Review runs, scored 49.2% at $0.102.
| Lane | Model | F1 | Rank among models | Time (p50) | $ / review |
|---|---|---|---|---|---|
| Code Review benchmark | Muse Spark 1.3 | 56.6% | 1 of 5 | 3m 18s | $0.424 |
| Security Review benchmark | Muse Spark 1.3 | 44.4% | 5 of 7 | 2m 30s | $0.258 |
| AI Architecture benchmark | Muse Spark 1.3 | 28.6% | 7 of 8 | 2m 04s | $0.193 |
| Log & Errors benchmark | Muse Spark 1.3 | 43.6% | 3 of 5 | 56s | $0.085 |
| Docs Review benchmark | Muse Spark 1.3 | 59.5% | 3 of 3 | 48s | $0.087 |
| API Performance benchmark | Muse Spark 1.3 | 56.0% | 1 of 6 | 42s | $0.063 |
| SOC 2 Review benchmark | Muse Spark 1.3 | 55.2% | 5 of 5 | 49s | $0.109 |
| CI/CD Performance benchmark | Muse Spark 1.3 | 44.4% | 5 of 5 | 49s | $0.034 |
Highlighted rows are what the agent runs today. Green: the best model on that lane.
FAQ
Muse Spark for review
Is Muse Spark good for code review?
For some reviews. Muse Spark 1.3 is the best model we ran on Code Review (F1 56.6%) and API Performance (F1 56.0%). On general code review it costs $0.424 a review; DeepSeek V4.1 Flash, which IonWarp’s Code Review runs, scored 49.2% at $0.102.
What does a Muse Spark review cost?
In model charges on the benchmark, from $0.034 a review on CI/CD Performance to $0.424 on Code Review. On IonWarp, model spend comes out of the review spend your plan includes.
Which Muse Spark models has IonWarp benchmarked?
Muse Spark 1.3, on 8 of the 8 lanes: Code Review, Security Review, AI Architecture, Log & Errors, Docs Review, API Performance, SOC 2 Review and CI/CD Performance.
Does IonWarp run Muse Spark?
Not by default: the agents start on DeepSeek V4.1 Flash. Code Review and API Performance retry a failed attempt on Muse Spark. An agent moves to another model only when IonWarpBench finds one that clears its quality bar for less, and every review names its model.

Every agent names its model.
Try for free14 days free, no card. Then Pro is $49 a month per workspace with 5 seats, or Max $149 with 10.