aquif-3.6-8B

Summary

aquif-3.6-8B is a hybrid reasoning model that automatically determines when and how deeply to think based on query complexity. Built on aquif-3.5-8B-Think with AutoThink RL data, it achieves 28% better token efficiency and 4% performance improvement across benchmarks.

Contents

• Key Features - Dynamic reasoning, efficiency gains, and smart resource allocation
• Performance - Benchmark results showing 4% average improvement
• Token Efficiency - 28% reduction in token usage
• Thinking Ratio - 12% reduction in thinking frequency
• Benchmark Highlights - Detailed results for AIME, LiveCodeBench, and GPQA Diamond
• Model Details - Architecture and specifications
• Usage - Code examples for implementation
• Previous Versions - Links to earlier models

Automatic Thinking

aquif-3.6-8B is a hybrid reasoning model that dynamically decides if and how much to think based on query complexity. Inspired by KAT-V1's approach of automatic thinking using AutoThink RL data on top of aquif-3.5-8B-Think, the model uses the following format:

<judge>
[analyzes whether to think or not]
</judge>

<think_on/off>
<think>
[thinking content]
</think>

<answer>
</answer>

This is the same format as KAT-V1-40B. Unlike something like DeepSeek-V3.1's toggleable reasoning that requires manual control (thinking_on/off), aquif-3.6's judge autonomously allocates reasoning depth - intelligently adapting its cognitive effort to each task automatically.

Key Features

• 🧠 Dynamic Reasoning: Automatically determines when and how deeply to think
• ⚡ 28% More Efficient: Significant token reduction while improving performance
• 📈 Better Performance: 4% average improvement across benchmarks
• 🎯 Smart Resource Allocation: 12% reduction in thinking ratio on average

Performance

Benchmark	aquif-3.6-8B	aquif-3.5-8B	Improvement
AIME 2025	82.5	81.4	+1%
LiveCodeBench	64.2	61.5	+4%
GPQA Diamond	71.0	66.8	+6%
Average	72.6	69.9	+4%

Token Efficiency

Benchmark	aquif-3.6-8B	aquif-3.5-8B	Reduction
AIME 2025	15,670	21,265	-26%
LiveCodeBench	13,240	19,460	-32%
GPQA Diamond	8,760	11,560	-24%
Average	12,557	17,428	-28%

Thinking Ratio

Benchmark	aquif-3.6-8B	aquif-3.5-8B	Reduction
AIME 2025	93.0%	100.0%	-7%
LiveCodeBench	82.0%	100.0%	-18%
GPQA Diamond	89.0%	100.0%	-11%
Average	88.0%	100.0%	-12%

Benchmark Highlights

• AIME 2025: 26% fewer tokens, +1% performance, -7% thinking ratio
• LiveCodeBench: 32% fewer tokens, +4% performance, -18% thinking ratio
• GPQA Diamond: 24% fewer tokens, +6% performance, -11% thinking ratio

Model Details

• Base Model: 8B parameters
• Architecture: Hybrid reasoning with dynamic thinking allocation
• Context Length: 40K tokens
• License: Apache 2.0

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "aquif-ai/aquif-3.6-8B"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto"
)

messages = [
    {"role": "user", "content": "Solve this problem: What is the sum of all prime numbers between 1 and 100?"}
]

input_ids = tokenizer.apply_chat_template(
    messages,
    add_generation_prompt=True,
    return_tensors="pt"
).to(model.device)

outputs = model.generate(
    input_ids,
    max_new_tokens=2048,
    temperature=0.7,
    do_sample=True
)

response = tokenizer.decode(outputs[0][input_ids.shape[-1]:], skip_special_tokens=True)
print(response)

Previous Versions

• aquif-3.5-8B-Think

---

Built by aquif-ai