# Laya

Decision models · checked 2026-10-14 · https://dbggr.dev/models/laya/

Laya is an open-weight decision model from Convai Innovations (Apache 2.0, 421M parameters): it answers typed questions, such as yes or no, pick one or pick several, with probabilities instead of writing text. It's currently free to use (Vercel AI Gateway). It scores 32.5 (self-hosted; seconds are the board's estimate; outside the board's latency cap) on the independent JevBench Capability Score (board v1.7.46), where Jev, the reference, scores 77.1.

- Made by: Convai Innovations
- Weights: Open, Apache 2.0
- Size: 421M
- Where: Vercel AI Gateway (free until 31 Oct)
- Images: No, text only
- Context: 512–8,192
- Fine-tuning: Notebook provided
- Docs: https://huggingface.co/convaiinnovations/laya

## Laya price

US dollars per million input tokens. Decision models don't write text, so none of the hosted ones charge for output.

| Where | Input | Output | Notes | Checked |
|---|---|---|---|---|
| Vercel AI Gateway (free until 31 Oct) | Free | Free | Free on Vercel's AI Gateway until 31 October. | [2026-10-14](https://vercel.com/changelog/laya-decision-model-now-available-on-ai-gateway-free-through-october-31) |

## Laya benchmarks

| Benchmark | Result | What it measures | Reported by |
|---|---|---|---|
| JevBench v1.7.46 | 32.5 | Capability Score: accuracy and calibration on typed decisions, with Jev as the reference system | [Independent](https://benchmarkheaven.com/jev-models) |
| JevBench time | 1.80 s | Seconds per decision: measured end to end for hosted APIs, the board's estimate for self-hosted models | [Independent](https://benchmarkheaven.com/jev-models) |
| Reported speed | 33 ms on a T4 GPU | Latency on the maker's own hardware | Convai Innovations |

Makers' speeds are measured on their own hardware and aren't comparable with each other or with JevBench's timings.

## Decision models like Laya

| Model | Input / M | JevBench | Weights |
|---|---|---|---|
| Laya | Free | 32.5 | Open |
| [Jev 1.13](https://dbggr.dev/models/jev/) | $0.042 | 77.1 | Closed |
| [Clef](https://dbggr.dev/models/clef/) | $0.24 | 75.3 | Open |
| [d1](https://dbggr.dev/models/d1/) | $0.04 | 74.4 | Closed |
| [OpenAI Decisions (GPT-6 Luna)](https://dbggr.dev/models/openai-decisions/) | $0.10 | 73.5 | Closed |
| [Microsoft-Decision-1](https://dbggr.dev/models/microsoft-decision-1/) | $0.042 | 70.8 | Closed |
| [Clef-flash](https://dbggr.dev/models/clef-flash/) | $0.038 | 63.8 | Open |
| [Clef-omni](https://dbggr.dev/models/clef-omni/) | $0.15 | 58.3 | Open |
| [d1-3B](https://dbggr.dev/models/d1-3b/) | Self-host | 49.5 | Open |
| [d1-omni-600M (experimental)](https://dbggr.dev/models/d1-omni-600m/) | Self-host | 32.8 | Open |
| [CLM-8B](https://dbggr.dev/models/clm-8b/) | Self-host | 20.1 | Open |
| [Strands Decider 2B](https://dbggr.dev/models/strands-decider-2b/) | Self-host | – | Open |

## FAQ

### How much does Laya cost?

Laya costs nothing at Vercel AI Gateway (free until 31 Oct). Free on Vercel's AI Gateway until 31 October. Checked 14 October 2026.

### Is Laya open weight?

Yes. Convai Innovations publishes the weights under Apache 2.0, so you can run and fine-tune it yourself.

### How fast is Laya?

Convai Innovations reports 33 ms on a T4 GPU, measured on its own hardware. JevBench gives 1.80 seconds per decision, its estimate for self-hosting.

### How does Laya compare with Jev?

On JevBench's Capability Score (board v1.7.46), Laya scores 32.5 against Jev's 77.1. Jev costs $0.042 per million input tokens; Laya costs Free. JevBench is a one-person project that uses Jev as its reference, which favours cheap hosted models.