RSI-Jev v6.1 VL 27B (effort auto)
RSI-Jev v6.1 VL 27B (effort auto) by shgao is a decision model: it answers typed questions with a probability for each allowed answer instead of generating text. It scores 58.7 on the Jev Decision Index (rank 15).
Pricing
No hosted price is listed. The weights are open: the cost is your own compute.
Specifications
- Provider
- shgao
- Parameters
- 28B
- Released
- Oct 2026
- Open weights
- Yes
- Question types
- Choice (one of N options), Score (ordered rubric), Yes / no
- Mechanism
- Decoder LM with a decision readout
- Backbone
- Qwen/Qwen3.8-27B
- Jev-compatible request
- Yes
- Runs on CPU
- No
- License
- apache-2.0
- Hugging Face
- shgao/rsi-jev-v6.1-vl-27b
Benchmarks
- Jev Decision IndexECE 0.077
- 58.7
- Jev Decision Index — VisionECE 0.037
- 64.0
Jev Decision Index: public 65.6, ECE 0.077, Brier 0.327. Lower ECE means its probabilities can be taken at face value.
Speed & latency
Compare RSI-Jev v6.1 VL 27B (effort auto) with…
FAQ
How much does RSI-Jev v6.1 VL 27B (effort auto) cost?
RSI-Jev v6.1 VL 27B (effort auto) has no hosted price listed; its weights are open, so the cost is the compute you run it on.
What questions can RSI-Jev v6.1 VL 27B (effort auto) answer?
RSI-Jev v6.1 VL 27B (effort auto) answers choice (one of n options), score (ordered rubric), yes / no questions, returning a probability for each allowed answer.
Is RSI-Jev v6.1 VL 27B (effort auto) well calibrated?
On the Jev Decision Index its expected calibration error (ECE) is 0.077; below about 0.05, a stated 80% confidence is right roughly 80% of the time.
Is RSI-Jev v6.1 VL 27B (effort auto) open weight, and can it run on a CPU?
Yes, RSI-Jev v6.1 VL 27B (effort auto) has open weights. It needs a GPU to run at useful speed.
Data as of 2026-10-14. Source: Metatext decision-model catalogue, 2026-10-14 (Jev Decision Index, Benchmark Heaven JevBench, Hugging Face). Pricing is per 1M tokens (USD); confirm with the provider before production use.