Ollama · kev

# Kev Kev is a family of text decision models from [Jared Palmer](https://huggingface.co/jaredpalmer). It reads a state and one or more typed questions, then scores the supplied answers in a forward …

ollama/kev

# Kev

Kev is a family of text decision models from [Jared Palmer](https://huggingface.co/jaredpalmer). It reads a state and one or more typed questions, then scores the supplied answers in a forward pass. Use Ollama's local `/v1/systemone` API for classification, yes/no decisions, and ordered scores. Kev does not generate text or accept images.

**Requires Ollama 0.40.4 or later.** These tags use the Apple Silicon MLX runner and advertise the `decision` capability.

## Choose a model

`kev` and `kev:latest` select **`kev:9b-nvfp4`**. It is the default balance of quality, response time, and memory on the tested M5 Max. The 27B variants are the quality choices; the 0.8B variants fit tighter memory budgets. All four publisher size checkpoints are available as BF16 tags. Lower-memory formats are additional choices, not replacements for those checkpoints.

| Tag | Format | Public JevBench correct | Hard subset correct | M5 Max resident memory¹ | | --- | --- | ---: | ---: | ---: | | `kev:0.8b-mlx-bf16` | BF16 | 146/231 | 39/111 | 3.28 GiB | | `kev:0.8b-mxfp8` | MXFP8 | 146/231 | 39/111 | 2.84 GiB | | `kev:4b-mlx-bf16` | BF16 | 175/231 | 60/111 | 12.47 GiB | | `kev:9b-mlx-bf16` | BF16 | 184/231 | 66/111 | 19.59 GiB | | **`kev:9b-nvfp4`** | mixed NVFP4/MXFP8 | **183/231** | **68/111** | **11.16 GiB** | | `kev:27b-mlx-bf16` | BF16 | 202/231 | 83/111 | 55.56 GiB | | `kev:27b-nvfp4` | mixed NVFP4/MXFP8 | 205/231 | 86/111 | 25.56 GiB |

¹ `/api/ps` resident size after a three-question workload on one Apple M5 Max with 128 GB unified memory. It is not a peak-memory requirement. All seven JevBench rows are final-runtime local results from the same 231 public tasks in JevBench v1.2.16. Each tag returned 231 valid responses. The results do not include the 303 unavailable sealed or judge tasks and are not an official leaderboard score. Quantization changed some individual answers and probabilities even where aggregate accuracy was close.

The 9B NVFP4 default answered 68/111 hard public tasks while using less observed resident memory than 4B BF16 (60/111, 12.47 GiB) or 9B BF16 (66/111, 19.59 GiB). The 27B NVFP4 tag answered 86/111 hard tasks and used 25.56 GiB resident, so it is the stronger choice when the extra memory and response time are acceptable. These are results from one public set, not guarantees for a different task. The default follows the [Tev1](https://ollama.com/library/tev1/tags), [Nimble](https://ollama.com/library/nimble/tags), and [Clef Flash](https://ollama.com/library/clef-flash/tags) pattern of selecting a capable compact variant while retaining explicit larger or source-precision tags.

## Try it

Pull a tag, then send a System One request to the local server:

```sh ollama pull kev

curl http://localhost:11434/v1/systemone \ -H 'Content-Type: application/json' \ -d '{ "model": "kev", "state": "The package arrived with a cracked screen.", "questions": { "route": { "type": "choice", "instructions": "Which team should handle this message?", "criteria": { "returns": "Damaged items and returns", "shipping": "Delivery and tracking", "billing": "Payments and charges" } } } }' ```

The `route` answer contains a `choice`, probabilities for each option, and a confidence value. Kev also supports `noul` questions (probability of true) and `score` questions (an expected level over ordered criteria). See the [System One guide](https://docs.ollama.com/capabilities/decision) for response shapes and more examples.

For Kev, `choice` accepts 1–255 named options and `score` accepts 1–255 ordered levels. Descriptions may be JSON values. Multiple questions can share a state in one request. Reported probabilities and confidence are rounded to four decimal places; confidence describes how concentrated the choices are, not the chance that the answer is correct. The tag's `num_ctx` is 73,728 tokens, but a large accepted input does not imply reliable accuracy at that length.

## Use and limitations

These models are designed for bounded decisions where the relevant evidence and permitted answers are supplied. Add an explicit `none` option when no listed answer may fit. Check output quality and probability thresholds on labeled examples from your own task, especially after changing model size or quantization. The 0.8B models are best suited to simpler decisions; in the local public JevBench run they got only 39/111 hard tasks correct.

The publisher's evaluations focus on English. Accuracy can fall on unfamiliar domains, long inputs, reordered options, and tasks that need facts absent from the state. A fitted temperature does not make probabilities universally calibrated. Do not use these models as the sole control for consequential decisions. [Kev 4B](https://huggingface.co/jaredpalmer/kev-4b/blob/6cfce5c2fa4b4bd64026336ab649c5ca78857d52/README.md) and [Kev 27B](https://huggingface.co/jaredpalmer/kev-27b/blob/af0e6d551bdc2cc724f3e9d7a8bee1cd4fb8f7bf/README.md) document the publisher's separate held-out and long-document evaluations and their limits.

## Source and license

The source checkpoints are [0.8B](https://huggingface.co/jaredpalmer/kev-0.8b/tree/bf75a6a8848ea6960ff2ed108d9ed44c2941174f), [4B](https://huggingface.co/jaredpalmer/kev-4b/tree/6cfce5c2fa4b4bd64026336ab649c5ca78857d52), [9B](https://huggingface.co/jaredpalmer/kev-9b/tree/db029f08b290afd9fee4aa4bbcd9ae48602d1eb0), and [27B](https://huggingface.co/jaredpalmer/kev-27b/tree/af0e6d551bdc2cc724f3e9d7a8bee1cd4fb8f7bf). The 0.8B, 4B, and 9B Ollama weights merge the publisher's adapters with their identified Qwen3.5 bases; 27B uses the publisher's full text checkpoint and its selected Qwen3.8 tokenizer. The pointer head remains FP32 in every format. The publisher and base-model licenses are Apache-2.0; the applicable license texts are included in each tag.

Specifications & capabilities

Published model ID
kev

Deployments & artifacts

1–16 / 16

ollama

source id
ollama
execution mode
local
parameter size
752.92M
parameter count
752920000
parameter count is approximate
true
quantization level
BF16
context window
256000
context window is approximate
true
artifact size bytes
1517821684
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter count is approximate
false
quantization level
BF16
context window
256000
context window is approximate
true
artifact size bytes
1519780654
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter count is approximate
false
quantization level
mxfp8
context window
256000
context window is approximate
true
artifact size bytes
1038314180
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
752.92M
parameter count
752920000
parameter count is approximate
true
quantization level
Q8_0
context window
256000
context window is approximate
true
artifact size bytes
812429044
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter count is approximate
false
quantization level
BF16
context window
256000
context window is approximate
true
artifact size bytes
51272636349
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter count is approximate
false
quantization level
nvfp4
context window
256000
context window is approximate
true
artifact size bytes
19055202385
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
4.2B
parameter count
4200000000
parameter count is approximate
true
quantization level
BF16
context window
256000
context window is approximate
true
artifact size bytes
8427043444
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter count is approximate
false
quantization level
BF16
context window
256000
context window is approximate
true
artifact size bytes
8429659459
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
4.2B
parameter count
4200000000
parameter count is approximate
true
quantization level
Q4_K_M
context window
256000
context window is approximate
true
artifact size bytes
2709570164
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
4.2B
parameter count
4200000000
parameter count is approximate
true
quantization level
Q8_0
context window
256000
context window is approximate
true
artifact size bytes
4483824244
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
9.0B
parameter count
9000000000
parameter count is approximate
true
quantization level
BF16
context window
256000
context window is approximate
true
artifact size bytes
17924919988
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter count is approximate
false
quantization level
BF16
context window
256000
context window is approximate
true
artifact size bytes
15894672165
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter count is approximate
false
quantization level
nvfp4
context window
256000
context window is approximate
true
artifact size bytes
6744833865
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
9.0B
parameter count
9000000000
parameter count is approximate
true
quantization level
Q4_K_M
context window
256000
context window is approximate
true
artifact size bytes
5630317236
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
9.0B
parameter count
9000000000
parameter count is approximate
true
quantization level
Q8_0
context window
256000
context window is approximate
true
artifact size bytes
9529758388
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
9.0B
parameter count
9000000000
parameter count is approximate
true
quantization level
Q4_K_M
context window
256000
context window is approximate
true
artifact size bytes
5630317236
updated at
2026-11-16T08:10:06.079865+00:00
metadata warning
Registry digest differs from the catalog hint. Both observations are retained; representation or tag revision may differ.

Original sources

1–2 / 2

Ollama · Model page ↗

Ollama · Catalog ↗

Source links

1–1 / 1

Explore more models

Models

Loading
EVIDENCE / COMPARISON

Benchmarks

Explore every benchmark

Loading

These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

Install TensorTrace