Invergent's decision model for text and images. A fine-tune of Gemma-4-26B-A4B (about 4B parameters active per token) that answers Choice, Noul and Score questions, with an optional thinking mode for harder questions.
A Gemma 4 12B fine-tune for typed decisions that also chats and reads images. Ships as GGUF for llama.cpp, holds a 64K context on a 16 GB GPU, and serves /v1/systemone next to /v1/chat/completions.
Decides
choice, score, noul, classify, route
choice, score, noul, classify, route
Architecture
rune
winnow
Fine-tuned from
google/gemma-4-26b-a4b-it
google/gemma-4-12b-it
License
apache-2.0
apache-2.0
Availability
Open weights + hosted API
Open weights
Hosted by
Invergent
—
Input price
—
—
Decision accuracy
—
85.7%
Calibration error
—
—
Valid action rate
—
—
Median latency
—
143 ms
p95 latency
—
—
Evaluation suite
Figures are from each model’s manifest; accuracy and latency are what the publishers report, on their own suites and hardware. Add a third model.
Questions
What is the difference between rune and winnow?
rune is from Surogate (Invergent) and winnow from EldanRing. rune has open weights and a hosted API; winnow has open weights you can download and run. Both answer choice, score, noul, classify and route questions. rune reads up to 262K tokens of state, against 64K tokens for winnow. winnow is the smaller model, at 12B parameters to 26B.
Which is more accurate, rune or winnow?
Only winnow publishes an accuracy figure (85.7% on JevBench public subset (231 items), Q8_0); rune does not, so there is no comparison to make without your own test.
Which is cheaper, rune or winnow?
rune: Hosted, price not published, or free to self-host. winnow: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run rune or winnow locally?
Yes, both: systemone pull surogate/rune and systemone pull eldanring/winnow download the weights.
