German startup's Endeavor 1.1 claims to beat OpenAI and Anthropic, met with grift accusations
teortaxesTex · x · 2026-11-03
German company Flora Labs released Endeavor 1.1, claiming reasoning, coding and agentic capabilities rivaling top OpenAI and Anthropic models, with EU GPU inference options and enterprise support — its third model this year. The benchmark screenshot looks suspicious: HumanEval 99.9, AIME 2026 99.9, GPQA 94.4, allegedly ahead of GPT-5.6 Sol, Claude Fable 5, Kimi K3 and Nemotron 3 Ultra. Prominent AI commentator teortaxesTex openly questioned what kind of "grift" this is. Unknown teams posting sweep-the-board scores are a classic red flag for benchmark contamination or marketing hype.
More from Fun
- Flower Labs' YC-backed federated learning site mocked as AI-generated slop — teortaxesTex · 2026-11-03
- Claude Code replicates Doppler-based location leak: pinpointing a webcam from audio alone — flngr · 2026-11-03
- Hugging Face dev open-sources the model he uses every day — lhoestq · 2026-11-03
- Google AI Mode keeps leaking image carousel metadata as plain text — One-Nobody639 · 2026-11-03
- Training neural nets inside LLMs: XOR MLP goes 53% to 98.8% in 40 steps on simulated QwenGPUs — zeeshanp_ · 2026-11-03
- Tiny latent diffusion model trained inside LLMs generates MNIST digits, 76% recognized — zeeshanp_ · 2026-11-03