REAX launches open decision model REAX 1 and miner competitions
airesearch12 · x · 2026-10-21
REAX Labs released REAX 1, an open decision model fine-tuned from Perplexity Decider v1.2, claiming +2.4 points over its base with wins across all 6 test draws and 10 areas. A public competition opens to let anyone beat it for rewards. To keep JevBench independent, REAX won't list its own models there, submitting instead to third-party benchmarks like the HF Decision Index and Artificial Analysis Decision Index. Competitions run on subnet 92 with demand-forecasting (baseline 27.32) and risk-assessment (baseline 28.54) tracks, scored on sealed unseen items; emissions split: 50% inference miners, 20% engine, 15% each fine-tune track.
More from Venture
- Inside the SEO community drama: long-tail keyword arbitrage and thin-shell tool sites — gefei55 · 2026-10-21
- OpenAI invests in startup using brain implants to detect cancer, Wired reports — wiredmagazine · 2026-10-21
- Dev launches modelanalysis.si to track and calculate LLM API costs without spreadsheet hell — zackku · 2026-10-21
- Every top-10 economy now has a national AI model; 350+ launches in two years, IDC finds — perilli · 2026-10-21
- Anthropic's S1 is reportedly the first prospectus ever to list doomsday-scenario risks — StewartalsopIII · 2026-10-21
- Solo dev reproduces agent failures and shrinks them to minimal cases, but can't find customers — ibackstrom · 2026-10-21