Best Llama4 AI Tools & Models - Premium Llama4 News

AI News

Meta Chief AI Scientist Reveals: Fabricated Test Results Before Llama 4 Launch Causes Industry Shock!

The Llama series of AI models developed by Meta was once highly praised, but the release of Llama4 in April 2025 sparked a trust crisis. Meta claimed its benchmark test results were excellent, but developers found the actual performance to be far below the advertised level, leading to suspicions that Meta may have used improper methods during testing.

12.4k 8 hours ago

Meta Super Intelligence Lab Breaks RAG Technology Bottleneck: The REFRAG Framework Boosts Inference Speed by 30 Times

Meta Super Intelligence Lab introduced the REFRAG technology, which improves the inference speed of large language models in retrieval-augmented generation tasks by more than 30 times. This breakthrough result was published in a related paper and profoundly transforms the way AI models operate. The lab was established in California in June this year, stemming from Zuckerberg's emphasis on the Llama4 model.

10.6k 2 days ago

Meta Super Intelligence Lab Breaks RAG Technology Bottleneck: The REFRAG Framework Boosts Inference Speed by 30 Times

DuckDuckGo Subscription Service Adds AI Features, $9.99 per Month to Access Top Models

DuckDuckGo's subscription service now includes free AI chat via Duck.ai, offering models like Claude3.5Haiku and Llama4Scout, alongside existing VPN and privacy features.....

11.3k 48 minutes ago

DuckDuckGo Subscription Service Adds AI Features, $9.99 per Month to Access Top Models

AMD Makes Major Upgrade! Ryzen AI Max+ 395 Makes It Possible to Run a 128-Billion-Parameter Model Locally!

AMD Ryzen AI Max+395 upgrades to support 128B-parameter local AI models, requiring 128GB RAM (96GB VRAM). Runs 109B-parameter models like Meta Llama4Sout at 15 tokens/sec with 256K context length. MoE architecture enables efficient processing. 128GB AI workstations now priced at ~$1,800.....

10.2k 8 hours ago

Models

Qianfan-QI-VL

Baidu

Input tokens/M

Output tokens/M

Context Length

QianfanHuijin-8B

Baidu

Input tokens/M

Output tokens/M

Context Length

QianfanHuijin-70B

Baidu

Input tokens/M

Output tokens/M

Context Length

Qwen3-8B

Alibaba

$0.5

Input tokens/M

Output tokens/M

Context Length

Qianfan-Llama-VL-8B

Baidu

Input tokens/M

Output tokens/M

Context Length

Gemma 3 27B

Google

$0.7

Input tokens/M

$1.4

Output tokens/M

131

Context Length

Hunyuan-Large

Tencent

Input tokens/M

$12

Output tokens/M

Context Length

DeepSeek-R1-Distill-Llama-8B

Deepseek

Input tokens/M

Output tokens/M

Context Length

Gemma 2 9B

Google

Input tokens/M

Output tokens/M

Context Length

Gemma 2 27B

Google

Input tokens/M

Output tokens/M

Context Length

Yi-6B-Chat

01-ai

Input tokens/M

Output tokens/M

Context Length

Yi-34B-200K

01-ai

Input tokens/M

Output tokens/M

200

Context Length

Yi-34B

01-ai

Input tokens/M

Output tokens/M

Context Length

Baichuan2-7B-Base

Baichuan

Input tokens/M

Output tokens/M

Context Length

Doubao-1.5-pro-256k

Bytedance

Input tokens/M

Output tokens/M

256

Context Length

Empowering the future, your artificial intelligence solution think tank

English 简体中文繁體中文にほんご

FirendLinks:

AI Newsletters AI Tools MCP Servers AI News AIBase LLM Leaderboard AI Ranking

Business Cooperation Site Map

AI News

Meta Chief AI Scientist Reveals: Fabricated Test Results Before Llama 4 Launch Causes Industry Shock!

Meta Super Intelligence Lab Breaks RAG Technology Bottleneck: The REFRAG Framework Boosts Inference Speed by 30 Times

DuckDuckGo Subscription Service Adds AI Features, $9.99 per Month to Access Top Models

AMD Makes Major Upgrade! Ryzen AI Max+ 395 Makes It Possible to Run a 128-Billion-Parameter Model Locally!

Models

Qianfan-QI-VL

QianfanHuijin-8B

QianfanHuijin-70B

Qwen3-8B

Qianfan-Llama-VL-8B

Gemma 3 27B

Hunyuan-Large

DeepSeek-R1-Distill-Llama-8B

Gemma 2 9B

Gemma 2 27B

Yi-6B-Chat

Yi-34B-200K

Yi-34B

Baichuan2-7B-Base

Doubao-1.5-pro-256k

Llama4Some SOVL 4x8B L3 V1