All coverage under this tag.
The gap between open models like Llama and DeepSeek and closed models like GPT and Claude keeps narrowing. A calm framework across cost, privacy, performance and ecosystem.
On July 27, 2026, Anthropic CEO Dario Amodei clarified the company has never advocated banning open-weight models, days after an Nvidia-led open letter urged against 'premature restrictions.'
30M free tokens a day, a zero-setup cloud dev environment, open-source and self-hostable — MonkeyCode is becoming the go-to choice for both individuals and enterprises adopting AI coding, already serving 100,000+ developers.
Competition in open-source LLMs is fierce, with several companies releasing new-generation open models.
Moonshot AI released and open-sourced Kimi K2.5 on January 27, 2026, adding large-scale visual data atop the K2 trillion-parameter base.
Black Forest Labs released FLUX.2 in November 2025 with multi-reference editing; FLUX.2-dev open weights landed on Hugging Face.
OpenAI released the gpt-oss-120b and 20b open-weight reasoning models on August 5, 2025 — its first open weights since GPT-2.
Moonshot AI unveils Kimi K2 with 1 trillion total parameters — the largest open model at launch — excelling at agentic and coding tasks.
Alibaba releases the Qwen 3 series; the flagship Qwen3-235B-A22B rivals top models like DeepSeek-R1 and o1 on major benchmarks.
Meta released Llama 4 Scout and Maverick on April 5, 2025 — its first natively multimodal open-weight models, with Scout supporting a 10M-token context.
DeepSeek open-sourced its R1 reasoning model on January 20, 2025; a week later it triggered an AI stock rout, with Nvidia plunging 17% in a day.
DeepSeek released V3 on December 26, 2024: 671B total / 37B active parameters, trained for a fraction of closed-model budgets, open and commercial-ready.
Stability AI released the SD 3.5 family in October 2024, but with FLUX rising fast, its dominance in open image generation is under challenge.