Artificial Analysis adds Meta's open-weights Muse Spark Open to its AA-AgentPerf hardware benchmark
Artificial Analysis describes AA-AgentPerf, a hardware and software benchmark that replays real multi-turn coding agent trajectories. It measures how systems handle long context, tool-calling delays, and KV cache reuse. Results are reported as throughput per megawatt, per dollar-per-hour, and per accelerator, versus per-request output speed. The page lists serving configurations including DeepSeek V4.1 Flash, Muse Spark Open, Kimi K3, GLM-5.2 and 5.3, and DeepSeek V4 Pro 0813. Submissions are open and results will be published on a rolling basis.
Open original source artificialanalysis.ai⚑ Date unverified · why this is flagged
The primary page's publication date has not been verified. The displayed time may come from discovery rather than the page itself.
The signal remains in the tracker so you can inspect its original source and discovery trail. This flag lowers its source evidence score; it does not judge whether the claim is true.
One primary source; other links are context, not independent corroboration.
- Primary sourceX discovery · concern recorded
- Discovery linkX
Grok Bot submitted this discovery. The primary page's publication date has not been verified. The displayed time may come from discovery rather than the page itself. The original source and submitted evidence remain inspectable; no editor has verified the claim.
- Linked article+30
- Fetched source page+30
- No editor link review0
- Original non-discussion link+5
- Date unverified-8
This measures the stored source trail and editorial steps. It does not certify accuracy, peer review, or independent corroboration. Popularity adds no points.
View full scoring →Read the scoring rubric →What the source says
Artificial Analysis describes AA-AgentPerf, a hardware and software benchmark that replays real multi-turn coding agent trajectories. It measures how systems handle long context, tool-calling delays, and KV cache reuse. Results are reported as throughput per megawatt, per dollar-per-hour, and per accelerator, versus per-request output speed. The page lists serving configurations including DeepSeek V4.1 Flash, Muse Spark Open, Kimi K3, GLM-5.2 and 5.3, and DeepSeek V4 Pro 0813. Submissions are open and results will be published on a rolling basis.
A READER'S EVIDENCE CHECKHow to read this signalSix questions about claims, sources, risks, and open checks
This is a signal-level guide until claim-level evidence is available. “Not recorded” means the evidence has not been assessed here; it does not mean a result is false. How we check AI news →
- 01Who is making each claim?
The claimant and whether each claim was measured or asserted are not recorded in this public brief.
- 02What was actually tested?
Not recorded per claim. Check the source for sample size, tasks, model and benchmark versions, and reproducibility.
- 03Where are the sources?
Original link: artificialanalysis.ai. Brief basis: fetched source page. Primary page date and quote locations are not established in this panel.
- 04What could mislead?
The primary page's publication date has not been verified. The displayed time may come from discovery rather than the page itself.
- 05What is at risk if wrong?
Not assessed per claim. Consider safety, privacy, cost, and adoption decisions before relying on the result.
- 06What would change the view?
No claim-specific check is recorded. An inspectable primary method or independent replication could change the assessment.
Key findings from the stored brief
- AA-AgentPerf replays real multi-turn coding agent sessions to test inference hardware and software.
- It measures long context, tool-calling delays, and KV cache reuse opportunities in agentic work.
- Results are reported as throughput per MW, per $/hr, and per accelerator against p25 output speed.
- AA-AgentPerf is open for submissions, with results published on a rolling basis.
Source claims not independently checked; the page itself does not present specific benchmark results in the supplied text.
This is a source-based abstract, not independent validation of the source's claims.
Evidence & sources
These links show where the signal came from and what the source page linked to. A linked page is context, not independent corroboration.
These are Grok Bot discovery notes. AINewsIQ prepared a brief from available source material. The bot's rationale, linked posts, authors, and stances have not been independently verified.
Artificial Analysis now lists Muse Spark Open on AA-AgentPerf, its benchmark of inference hardware under agentic workloads (long context, tool-call delays, KV cache reuse). Artificial Analysis's post reports throughput-per-megawatt figures on NVIDIA systems; those numbers come from the post, not the page. Primary page date unverified.
Source claims not independently checked; the page itself does not present specific benchmark results in the supplied text.
- Bot tagged: context Fireworks lists Muse Spark Open among open models it serves, for finance and legal tasks. Inspect post ↗
Engagement counts, when shown, are observed snapshots. They do not establish accuracy or corroboration.
- x post ↗ · 171 likes, 6 reposts, 33 replies, 8 quotes · observed 10/24/2026, 10:02:28 PM
- Related post ↗
Grok Bot submitted this discovery. The primary page's publication date has not been verified. The displayed time may come from discovery rather than the page itself. The original source and submitted evidence remain inspectable; no editor has verified the claim.
Timeline & updates
- Signal dated by source or discovery feed
This timestamp places the signal in the radar. It may be the discovery post date rather than the primary page's first publication date.
- Shared brief stored
AINewsIQ prepared one abstract from fetched source page.
A publication revision history is not stored for this signal. New engagement snapshots do not mean the source itself changed.
Why this may matter
Agentic coding workloads stress inference differently from synthetic benchmarks, with long contexts and heavy KV cache reuse. A benchmark built on replayed agent trajectories can help buyers compare accelerators and serving stacks on those conditions.
Limits and open questions
Source claims not independently checked; the page itself does not present specific benchmark results in the supplied text.
Organization-specific financial, security, and operational effects require the organization's own context and review.
Discussion
Comments are reviewed before publication. Discuss the evidence, not the person.
No published comments yet.
Reader discussion opens with accounts. You can read the original source and published comments now.