Today for AI
HOT RADAR
llmHEAT 15.6°

AI CURATED HOT STORY · 10/30/2026

Baseten Open-Sources joey-1: A 2B Decision Model for Zero-Token Routing

1 reports archived1 independent sourcesupdated 10/30/2026, 05:22:18
Synthesis & Latest Updates
1 Sources Cross-Validated

Baseten has open-sourced joey-1, a 2B-parameter decision model that outputs calibrated probabilities via a single prefill pass without generating any text tokens. Scoring 63.30 on the Jev Decision Index, it ranks as the top model under 4B parameters with a P50 latency of just 20ms, offering a highly efficient solution for agent routing and classification tasks.

LATEST/Baseten has open-sourced joey-1, a 2B-parameter decision model that outputs calibrated probabilities via a single prefill pass without generating any text tokens. Scoring 63.30 on the Jev Decision Index, it ranks as the top model under 4B parameters with a P50 latency of just 20ms, offering a highly efficient solution for agent routing and classification tasks.

Hourly heat curve

2 fully observed hours
Fewer than 3 fully observed hours; trend not plotted yet.

Coverage timeline

Total 1 reports · Latest first
  1. Marktechpost (@Marktechpost)T2·78 pts

    Baseten Open-Sources joey-1: A 2B Decision Model for Zero-Token Routing

    • joey-1 utilizes a 'state + typed questions' paradigm to return probability distributions via a single prefill, eliminating text parsing overhead and generation costs.
    • Scores 63.30 on the Jev Decision Index, ranking #1 among models under 4B parameters and outperforming Perplexity Decider Flash v1.2 at the same price point.
    • Delivers ultra-low latency with a P50 of 20ms for 2,000-token requests, approximately 2.5x faster than its 9B sibling model, elephant-1.
    • Released under MIT license on Hugging Face, supporting self-hosting on a single CUDA GPU with input costs as low as $0.01 per million tokens.