Baseten has released joey-1, a 2B model from its in-house research lab, Base Labs. The joey-1 decision model returns a label and a calibrated probability instead of generated text. It targets high-volume routing, classification and guardrail checks. Baseten prices it at $0.01 per million input tokens, with open weights on Hugging Face under MIT.
TL;DR
- Size: 2B parameters, post-trained from Qwen3.5-2B. 16,384-token max sequence (Baseten says 262k on dedicated deployments).
- Runs on: Baseten Model APIs, or self-hosted on 1 CUDA GPU (VRAM not disclosed). No official quantized builds listed.
- Performance: Top model under 4B on the Jev Decision Index 0.4 public split, per Baseten.
- Best: 63.30 on Jev Decision Index 0.4 (public split), ranked #37 of 230 models.
- Worst: Trails its 9B sibling elephant-1 by 8.94 points on the same split (63.30 vs 72.24).
- Bottom line: Best: 20 ms P50 decisions for $0.01 per 1M tokens. Worst: less general knowledge than its larger siblings.
What is joey-1?
joey-1 is a 2B decision model from Baseten’s Base Labs that answers typed questions in 1 forward pass. You send a state (text or images) and 1 or more questions. The model returns a probability for every option of every question. It does not write free-form text that your application must parse.
It supports 3 question types:
- choice: pick 1 named option, such as which team owns a ticket.
- yes/no (
noulin the API): returns the probability of yes. - score: ordered levels, lowest first, such as urgency.
joey-1 is the third member of a family. Baseten shipped mammoth-1 (27B) and elephant-1 (9B) last week. All 3 share training data and the /v1/systemone API, so switching is a 1-line change.
How does joey-1 work?
Base Labs post-trained Qwen3.5-2B with a full fine-tune of the language model. The vision encoder stays frozen. Full-attention layers use a block-bidirectional mask. All questions are answered from 1 shared-prefix prefill, so extra questions add little cost.
Training details from the model card:
- Data: 6.21M decision examples. Mix: 58.3% core decision skills, 16.1% safety, 16.1% judging and expert domains, 9.5% public data.
- Objective: cross-entropy plus a Brier calibration term, with a trained “I don’t know” option.
- Recipe: 1 epoch, 43,390 steps, peak learning rate 1e-5, bf16.
- Hardware: 2 nodes of 8 H100 GPUs.
Calibration lets you set a confidence threshold. Uncertain answers can escalate to elephant-1, mammoth-1 or a human.
‘;}).join(”);
requestAnimationFrame(function(){requestAnimationFrame(function(){el.querySelectorAll(‘.fl’).forEach(function(f){f.style.width=f.getAttribute(‘data-w’)+’%’;});});});}
function drawBench(k){cur=k;var d=D[k];bars(document.getElementById(‘mj-bars’),d.rows,d.max,function(v){return k===’jev’?v.toFixed(2):v.toFixed(1).replace(/.0$/,”)});document.getElementById(‘mj-bnote’).textContent=d.note;setTimeout(post,60);}
R.querySelectorAll(‘[data-b]’).forEach(function(b){b.onclick=function(){R.querySelectorAll(‘[data-b]’).forEach(function(x){x.classList.remove(‘on’)});b.classList.add(‘on’);drawBench(b.getAttribute(‘data-b’));};});
function drawLat(){bars(document.getElementById(‘mj-lat’),[[‘joey-1′,20,’me’],[‘elephant-1′,49,’sib’]],60,function(v){return v+’ ms’});}
/* calc */
var r=document.getElementById(‘mj-r’);
function fmt(x){return ‘$’+(x>=1000?x.toLocaleString(‘en-US’):x);}
function calc(){var n=+r.value;var tok=n*1e6*500/1e6;document.getElementById(‘mj-n’).textContent=n+’M’;
document.getElementById(‘mj-c1’).textContent=fmt(Math.round(tok*0.01));document.getElementById(‘mj-c2’).textContent=fmt(Math.round(tok*0.035));document.getElementById(‘mj-c3’).textContent=fmt(Math.round(tok*0.10));}
r.addEventListener(‘input’,calc);calc();
window.addEventListener(‘load’,post);window.addEventListener(‘resize’,post);setTimeout(post,300);
})();
