15.8B
- input
- $0.01
- output
- $0.00
- Context
- 16Ktokens
- Released
- 2w ago
- Size
- 2.7Bparameters
Cost/1M tokens
Baseten's research lab Base Labs released joey-1, a decision model post-trained from Qwen3.5-2B with the same data and recipe as elephant-1 and mammoth-1, adding a third size to the Base Labs decision model family. The launch announcement calls it the family's fastest and most affordable decision model, claiming a P50 end-to-end latency of 20 ms for an average 2,000-token request. It is built for high-volume, latency-sensitive calls that need a decision and confidence score rather than generated text.
- Decision model
- Open weights · 2.7B
- Accepts images
How it compares
Frequently asked questions
Running locally
Joey 1 needs an estimated 2.4 GB of VRAM at Q4, 3.8 GB at Q8 or 6.8 GB at FP16, derived from its 2.7B parameters. All figures are estimates, covering weights plus about 15% and 0.5 GB for a short context.
Yes. At FP16 (estimated 6.8 GB) Joey 1 fits on 24 GB, 16 GB, 12 GB and 8 GB GPUs.
Yes. At Q4 Joey 1 needs an estimated 2.4 GB, and the smallest common Mac memory size that fits is 8 GB, assuming macOS lets the GPU use about 75% of unified memory.
API and pricing
$0.01 per 1M input tokens and $0.00 per 1M output tokens, based on 1 paid provider tracked.
$0.0001 at the cheapest provider, BaseTen, for a request of 10,000 input tokens and 1,000 output tokens.
In the last day, 1 endpoint had uptime data and 1 ran at 99% uptime or better, with BaseTen best at 100.0%.
Privacy and data
BaseTen offers zero data retention (ZDR eligible or 0-day retention).
Yes. Joey 1 has open weights, so you can self-host it for full control of your data.
Capabilities and specs
The context window is 16K tokens (16,384) and the maximum output is 15K tokens (14,745).
Joey 1 reads text and images. Audio, video and PDFs are not listed as supported input modalities.
Quality and benchmarks
Matched on model kind, blended price and release date: Span-02 by Respan and Decider Flash V1.2 by Perplexity, both at $0.0075 per 1M tokens blended, and Span-01 by Respan at $0.015 per 1M tokens.
Span-02Joey 1 vs Span-02Decider Flash V1.2Joey 1 vs Decider Flash V1.2Span-01Joey 1 vs Span-01
Licensing and access
Joey 1 is open weights.
The Joey 1 weights can be downloaded from the Hugging Face repository baselabs/joey-1.
Basics
Joey 1 was made by Baseten and released on October 29, 2026.
No. The listed API price is $0.01 input and $0.00 output per 1M tokens, though the weights are open.
Generated from Vertical Tabs data, updated November 2, 2026.
Discussion
No comments yet
Be the first to share your thoughts.