Zouk 1
Our fast, light model. Zouk 1 runs on the same weights as Zouk 2 in a lighter setup: shorter replies and a shorter memory, with the same direct voice. Quick answers without the wait.
- Parameters
- 24B
- Context
- 24k tokens
- Longest reply
- ~600 words
- Memory
- 16 messages
- Thinking
- Visible
- Plans
- Pro, Max, Max 20x and the trial
Same weights as Zouk 2, lighter configuration.
What it can do.
Reasoning
Thinks before it answers, with the thinking there to open if you want it.
Short-form writing
Ideas, posts and replies, up to ~600 words at a time.
Instruction following
Sticks to the tone and rules you set for the chat.
How it's set up.
Zouk 1 next to our other models, as each one runs on Bleepless.
Longest reply (words)
Chat memory (messages)
Context window (tokens)
Benchmarks.
Base-model benchmarks (developer-reported)
- ARC-cGrade-school science, 0-shot91.29
- MMLU57 academic subjects, 5-shot80.73
- GSM8KMath word problems, 5-shot80.73
- MBPPPython coding, pass@169.64
- MMLU ProHarder MMLU, 5-shot with reasoning54.37
- MATHCompetition math, 4-shot45.98
- GPQA MainGraduate-level science, 5-shot with reasoning34.37
Strengths
- Quick, to-the-point answers
- Replies up to ~600 words
- Remembers the last 16 messages, with a 24k-token context
- Same reasoning-trained weights as Zouk 2
Best for
- Quick answers
- Short replies and ideas
- Trying Bleepless out
Questions about Zouk 1.
What is Zouk 1?
Zouk 1 is the fast, light model from Bleepless AI. It runs the same 24-billion-parameter weights as Zouk 2 in a lighter setup, with replies up to ~600 words and a memory of the last 16 messages.
Is Zouk 1 uncensored?
Yes. Like Zouk 2, it's fine-tuned to answer rather than refuse, so you get a straight answer instead of a lecture. A few hard limits still apply, listed in our guidelines.
Which plans include Zouk 1?
Every plan, including the 7-day trial. If you need longer replies or more memory, Zouk 2 and Helios 1 are on the higher plans.
