Nanbeige4.1-3B

Nanbeige·Released 2026-02-11

Open Source3BApache 2.0

Compare View on HuggingFace

Quality Index

16.1

242nd of 440

Top 55%

Coding Index

8.9

278th of 350

Top 80%

Price/1M

$0.00

3rd cheapest

100% below median

Top 27%

Speed

0 tok/s

TTFT

0.00s

Market Position

Nanbeige4.1-3BMarket Average

Pricing

Input

$0.00

per 1M tokens

Output

$0.00

per 1M tokens

Blended

$0.00

per 1M tokens

Cheaper than 73% of models. Median price is $0.31/1M tokens.

Cost Calculator

Tokens per day1M

100K100M

Daily

$0.00

Monthly

$0.00

vs. Similar Models

Qwen3 VL 30B A3B InstructQ:0.0

$0.35

Qwen3 14B (Reasoning)Q:+0.1

$1.31

DeepSeek R1 Distill Llama 70BQ:-0.1

$0.88

Gemini 1.5 Pro (Sep '24)Q:-0.1

$0.00

Performance

0

tokens/sec

Faster than 0% of models

0.00

seconds

Faster than 61% of models

Market Median

46 tok/s

100% slower

Median TTFT

0.43s

100% faster

Speed Comparison

MiMo-V2-Pro

0 tok/s0%

GPT-5.2 (medium)

0 tok/s0%

mimo-v2-omni

0 tok/s0%

Benchmarks

MMLU-ProNot evaluated

GPQA Diamond

84.9%

HLE

10.0%

LiveCodeBenchNot evaluated

SciCode

26.6%

TerminalBench Hard

0.0%

MATH-500Not evaluated

AIMENot evaluated

AIME 2025Not evaluated

IFBench

35.4%

Long Context Recall

0.0%

Tau2

21.6%

Market AverageTop Score

Open Source

737.6K

Downloads (30d)

1.0K

Likes

apache-2.0

Permissive license allowing commercial use, modification, and distribution.

Hardware Requirements

Parameters3B

Est. VRAM (FP16)4-8 GB

Example GPURTX 3060 / M1

View on HuggingFace

Quick Compare

Similar Models

Qwen3 VL 30B A3B Instruct

Alibaba

Q: 16.1$0.35/1M131K ctx

Coding: +5.4

Qwen3 14B (Reasoning)

Alibaba

Q: 16.2$1.31/1M

Coding: +4.2

DeepSeek R1 Distill Llama 70B

DeepSeek

Q: 16.0$0.88/1M131K ctx

Ministral 3 14B

Mistral

Q: 16.0$0.20/1M

Hermes 4 - Llama-3.1 70B (Reasoning)

Nous Research

Q: 16.0$0.20/1M

Coding: +5.5

Gemini 1.5 Pro (Sep '24)

Google

Q: 16.0$0.00/1M

Coding: +14.7

Compare all 7 models