The story so far
Weights, every dated article, and repo commits on one rail.
18 repo commits since the weights landed
Teal is the weights (2026-09-24), grey ticks are repo commits.
Winnow-E4B
EldanRing Apache 2.0 out 31 days ago
Images and text in, text out.
Verified 2026-10-25 Rank #5730 in Decision models · score 47.5 method v1 Sources 3 Downloads 30d 58.1k · #370 of 619 tracked
Do people use it?58.1kdownloads, 30 days#370 of 619 we track · below the median+1 likes, downloads +13% this week12 likesHugging Face2026-10-24
Where it's used
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
No API provider serves it, so its downloads are people running it themselves: 1 quantized builds and 5 Spaces.
- As a service0no API provider listed
- On own hardware1quantized builds
- Built on10 fine-tunes · 1 adapters · 0 merges
- In apps5Hugging Face Spaces
Top Spaces running it: multimodalart/jev-decision-index ♥ 802
What people built on it
2 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
2 derivatives on the Hub, led by 1 quantized
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Reference
Everything else, folded: the full tables, every source.
Benchmarks
4 rows
Benchmarks 4 rows
measured third-party leaderboard rowauthor-reported number from the model card or eval files (4)No confidence intervals unless the source gave them; settings differ between rows, so compare within a benchmark, not across. Date = the benchmark's own date when it gives one, else when we read it.
| Benchmark | Value | Setting / notes | Reported by | Date |
|---|---|---|---|---|
| author-reported JevBench public subset, 231 questions | 80.52% percent | Winnow-E4B Q8 | model author | 2026-10-07 |
| author-reported Kev v9 additional, 390 questions | 55.90% percent | Winnow-E4B Q8 | model author | 2026-10-07 |
| author-reported Kev v9 clean, 1,046 questions | 72.66% percent | Winnow-E4B Q8 | model author | 2026-10-07 |
| author-reported Local typed decisions, 2,000 decisions | 72.30% percent | Winnow-E4B Q8 | model author | 2026-10-07 |
What the card says
2 cited sentences
Summary generated, cited
Winnow-E4B is an EldanRing fine-tune of the Gemma 4 E4B IT model designed for typed decisions, chat, and image input.[1] It supports image-text-to-text pipelines and can be used for tasks like routing requests or rating urgency.[2][3]
- “Winnow-E4B is a compact EldanRing fine-tune of Gemma 4 E4…”
- pipeline tag
- “routing requests, choosing actions, checking conditions, …”
Generated 2026-10-07 15:17 UTC by parasail-gemma-4-26b-a4b-it; each numeral points at the fact it rests on. Not a verdict.
Use-case evidence
3 use cases
What it is good at 3 use cases
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Facts
11 fields
Listed in Decision models · LLMs
Facts
- License: Apache 2.0 (friction: none). Permissive; patent grant. (license file, observed 2026-10-24)
- Created 2026-09-24, last modified 2026-10-13. (HF API, observed 2026-10-24)
| Pipeline tag | image-text-to-text |
|---|---|
| Library | gguf |
| MoE | no |
| License | Apache 2.0 (apache-2.0) |
| Hub snapshot | 2026-10-25 |
Decision facts
catalog
Decision-model catalog curated
| How it was made | LoRA/adapter fine-tune |
|---|---|
| Lineage bucket | F - fine-tune of an open model (public/synthetic data) |
| Lineage evidence | board file inspection (adapter/head/weight files) + HF base_model field |
| Base model | google/gemma-4-E4B |
| DI 0.2.1 score | 39.89 (independent (DI 0.2.1 board)) |
| ECE (board) | 0.0580 |
| Median latency | 45 ms (DI board, 1x RTX PRO 6000, single process) |
| Hosted price (input) | self-hosted/M |
| Source | https://huggingface.co/EldanRing/Winnow-E4B |
All variants
2 variants
Lineage
Base model: google/gemma-4-E4B-it
quantized: 1adapter: 1
Top variants (2 shown, ranked by downloads)
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| Txoka/Credence-v1-Gemma4-E4B | quantized | GGUF | — | — | apache-2.0 | Txoka | 0 | 484 |
| Txoka/Credence-v1-Gemma4-E4B-MiCA | adapter | — | — | — | apache-2.0 | Txoka | 0 | 0 |
Repository changelog
18 commits
Repository changelog 18 commits
| Date | Commit | SHA |
|---|---|---|
| 2026-10-13 | Add decision-model tag to model card | 17871fd |
| 2026-10-06 | Fix checksum and license links | ed89143 |
| 2026-10-06 | Publish approved Winnow card, usage and evidence | 804e129 |
| 2026-10-05 | Clarify target downloads and optional MTP assistant metadata | aabbd52 |
| 2026-10-04 | Fix checksum and license file links in model card | b29f0a4 |
| 2026-10-04 | Publish Winnow runtime guidance, measured results and model cards | 596563e |
| 2026-10-04 | Add verified NVFP4 and matching MTP assistant assets | 398eb4a |
| 2026-09-30 | Classify Winnow-E4B as image-text-to-text | 1b257e8 |
| 2026-09-24 | Update E4B comparison figure title | 734302f |
| 2026-09-24 | Add verified benchmark figures to model card | ba9554e |
| 2026-09-24 | Refocus Winnow-E4B model card on model and decision workflow | d1652cf |
| 2026-09-24 | Recommend Q8 profiles for 10 GB and 12 GB GPUs | 14c4224 |
| 2026-09-24 | Package E4B vision projector and verified 64K profile | a9b27ff |
| 2026-09-24 | Clarify BF16 validation scope | 7a24d1f |
| 2026-09-24 | Add evaluated step-438 BF16 GGUF | 22c8d8b |
| 2026-09-24 | Add E4B model card, evaluation, and quickstart | 7bd6326 |
| 2026-09-24 | Add Winnow-E4B Q8 model weights | 5793829 |
| 2026-09-24 | initial commit | db3f099 |
Similar models
3 models
Similar models
- llava-v1.5-7b same task (image-text-to-text)
- opt-125m closest by size and task
- gpt-j-6b permissive license
Sources
3 sources
Sources
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:43 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:17 UTC · Last verified 2026-10-25
Follow Winnow-E4B
Watch this model
- /feeds/models/EldanRing--Winnow-E4B.xml(RSS 2.0) · every change to Winnow-E4B
- /feeds/families/EldanRing--Winnow.xml(RSS 2.0) · every change across the Winnow family
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.