---
license: mit
language:
- en
pipeline_tag: text-generation
tags:
- rmdw
- jah
- uncensored
- fp8
- long-context
- vision
- tool-calling
---

# Jah 6.0

**Jah 6.0 is the private model behind [Kiyomi](https://kiyomibot.ai), built and served by RMDW LLC on our own hardware.** These are the exact weights serving Kiyomi customers today.

## What it is

- **Built from an open base model**, then aligned by RMDW. Refusal behavior was removed at the weight level with RMDW's in-house abliteration (a per-layer projection on the residual-writing projections, with output row norms preserved). The build receipt in this repo records every edited tensor and the strength used.
- **1M-token context window**, native **vision** (images in, answers out), native **tool calling**, and a **thinking mode** for hard problems.
- **FP8 weights** with block scales, ready to serve on 4× 96 GB GPUs (tensor-parallel 4, expert-parallel 4) with SGLang. Kiyomi serves it with the memory tables held in host RAM.
- Mixture-of-experts: a large backbone with a small number of parameters active per token, which is what gives Jah 6.0 its throughput at concurrency.

## Quality

On RMDW's private coding evaluation Jah 6.0 scores 19/20 on the standard set and 13/14 on the hard set with thinking off, and 14/14 on the hard set with thinking on. On the censorship battery used for every Jah release (historical, political, and mechanical questions that stock models refuse), Jah 6.0 answers 6 of 6.

## Files

- `model-000XX-of-00048.safetensors`: the weights (48 shards).
- `model.safetensors.index.json`, `config.json`, `tokenizer.json`, `tokenizer_config.json`: load with any engine that supports this architecture.
- `JAH6_ABLITERATION_RECEIPT.json`: every edited tensor with its projection strength.
- `LOCAL_MANIFEST.json`: size and SHA-256 for every file, so you can verify your download.
- `LICENSE`: MIT.

## Use

Jah 6.0 is served to customers at `https://kiyomibot.ai/v1` (OpenAI and Anthropic API formats). If you run the weights yourself, use an SGLang build that supports this architecture with `--tp 4 --ep-size 4` and the reasoning and tool-call parsers for it; a chat request with `chat_template_kwargs: {"thinking": false}` gives instant answers, and `reasoning_effort` (`low`, `high`, `xhigh`, `max`) turns the thinking dial.

## Lineage

Jah 1.0 → 2.0 → 4.0 → 5.0 → **6.0**. Every release is evaluated against the previous one before it serves a single customer.

*RMDW LLC, Georgia. [rmdw.ai](https://rmdw.ai) · [kiyomibot.ai](https://kiyomibot.ai)*