Founder claims Meta's Muse Spark 1.3 is benchmark-gamed, only GPT-5.6 Terra class in real use
bindureddy · x · 2026-09-03
Serial founder Bindu Reddy criticized Meta's Muse Spark 1.3, saying it "only looks good on benchmarks" and is in practice "just a GPT 5.6 Terra class model," alleging the benchmarks are being severely gamed. Unverified personal claim.
Related event: Founder Claims Meta Muse Spark 1.3 Benchmarks Are Misleading(2 posts)→
More from Models
- Hacker upgrades stolen account to $200 ChatGPT Pro with no fraud checks from OpenAI or bank — CucharitaDePalo · 2026-09-03
- Muse Spark 1.3 Update Gains Attention, Users Urged Not to Miss — D3VAUX · 2026-09-03
- 64GB Mac users debate best sub-40B MoE: is Qwen-3.8-35B worth the wait? — chibop1 · 2026-09-03
- xAI's mystery model 'Sol' spotted stealth-testing at 750 tps, rumored to drop soon — koltregaskes · 2026-09-03
- Microsoft's MAI-Transcribe-2 hits 2.0% WER at 411x real time for $1.67 per 1,000 minutes — ArtificialAnlys · 2026-09-03
- Claude Fable 5.1 scores 86.6% on SimpleBench, beating human baseline — Profanion · 2026-09-03