Skip to content
Only First-party#Open-source ecosystem

#Open-source ecosystem

Apr 14

Apr 14Tue
  1. Moonshot AI (Kimi) · new models on Hugging FaceAI score78

    Moonshot AI releases open-source Kimi K2.6 multimodal agentic model

    Moonshot AI released Kimi K2.6, an open-source native multimodal agentic model with 1T total and 32B activated parameters and a 256K context length. The model card reports benchmark results against GPT-5.4, Claude Opus 4.6, and Gemini 3.1 Pro across agentic, coding, reasoning, and vision tasks, and supports swarms of up to 300 sub-agents.

    AIWhy it matters: The model card gives specific agent swarm scale, context length, and benchmark comparisons against several frontier models, useful for judging its coding and agent capabilities.

Apr 8

Apr 8Wed
  1. MiniMax · new models on Hugging FaceAI score78

    MiniMax releases open-weight MiniMax-M2.7 with agent and coding gains

    MiniMax has released MiniMax-M2.7 on Hugging Face, describing it as its first model to participate in its own evolution. The source reports 56.22% on SWE-Pro, 46.3% on Toolathon, and 62.7% on MM ClawBench, and says an internal version autonomously optimized a programming scaffold over 100+ rounds for a 30% performance improvement.

    AIWhy it matters: The source ties its benchmark claims to a self-evolution process and a named comparison set, which helps readers weigh how the reported gains were achieved.

Apr 3

Apr 3Fri
  1. Z.ai (GLM) · new models on Hugging FaceAI score73

    Z.ai releases GLM-5.1, a flagship model for agentic engineering

    Z.ai has released GLM-5.1, its next-generation flagship model for agentic engineering, with stronger coding than GLM-5. The model is described as staying effective over longer agentic tasks, sustaining optimization over hundreds of rounds and thousands of tool calls. The release lists benchmark results including SWE-Bench Pro at 58.4 and Terminal-Bench 2.0 at 63.5, and local deployment is supported through SGLang, vLLM, xLLM, Transformers, and KTransformers.

    AIWhy it matters: The release gives benchmark tables against several rival models, letting readers compare GLM-5.1's coding and agentic results with GLM-5 and frontier systems.

Mar 31

Mar 31Tue
  1. Mistral AI · new models on Hugging FaceAI score76

    Mistral Medium 3.5 releases as a 128B dense merged model with vision

    Mistral AI released Mistral Medium 3.5, a dense 128B model with a 256k context window that handles instruction-following, reasoning, and coding in a single set of weights. It replaces Mistral Medium 3.1, Magistral, and Devstral 2, and reasoning effort is configurable per request. The model accepts text and image input and is released under a Modified MIT License that excludes companies with large revenue.

    AIWhy it matters: The release merges instruction, reasoning, and coding into one 128B model with per-request reasoning control, giving developers one set of weights to compare against separate specialized models.

  2. Alibaba NLP (Tongyi) · new models on Hugging FaceAI score26

    LaSER-Qwen3-8B: Alibaba NLP's 8B dense retriever with latent reasoning released on Hugging Face

    Alibaba NLP released LaSER-Qwen3-8B, an 8B-parameter dense retriever built on Qwen/Qwen3-8B that internalizes explicit reasoning into latent space through continuous latent thinking tokens. The model scores 29.3 nDCG@10 on the BRIGHT benchmark, ahead of the rewrite-then-retrieve pipeline's 28.1, and carries a 4096-dimension embedding with an 8192-token maximum sequence length. It is licensed under MIT and adds about 1.7× latency over standard single-pass dense retrievers.

Mar 24

Mar 24Tue
  1. ARC PrizeAI score70

    ARC Prize announces ARC-AGI-3, an interactive benchmark for frontier agents

    ARC Prize has released ARC-AGI-3, a set of hundreds of interactive, turn-based environments with thousands of game-style levels, with no instructions or stated goals. Humans score 100% while frontier AI scores 0.51%. ARC Prize 2026 offers over $2 million in prizes for open-source solutions to ARC-AGI-2 and ARC-AGI-3.

    AIWhy it matters: The benchmark's human versus frontier AI gap and its interactive design show how agent evaluation is shifting from instruction-following toward exploration and adaptation.

Mar 17

Mar 17Tue
  1. MiniMax BlogAI score63

    MiniMax M2.7 takes part in its own model and harness evolution

    MiniMax says M2.7 is its first model to deeply participate in its own evolution, building agent harnesses and running reinforcement learning experiment workflows. The post reports 56.22% on SWE-Pro, 55.6% on VIBE-Pro, 57.0% on Terminal Bench 2, and a 30% improvement on an internal evaluation set after more than 100 autonomous optimization rounds. It also states that M2.7 handles 30%-50% of its research team's workflow, though human researchers still make critical decisions.

    AIWhy it matters: The post ties M2.7's self-evolution claims to specific benchmark numbers and workflow details, helping readers judge how much of the iteration loop is autonomous.

  2. Apple · new models on Hugging FaceAI score46

    Apple releases SimpleSD-4B-instruct, a self-distilled Qwen code model

    Apple has released SimpleSD-4B-instruct on Hugging Face, a research checkpoint fine-tuned from Qwen3-4B-Instruct-2507 on its own sampled outputs to improve code generation. On LiveCodeBench, the model scores 41.5% pass@1 on LCBv6, up from the base model's 34.0%, and 45.7% pass@1 on LCBv5, up from 34.3%. The model is released under the Apple Machine Learning Research Model License and is intended for reproducibility rather than as an optimized Qwen release.

Jan 14

Jan 14Wed
  1. Black Forest Labs · new models on Hugging FaceAI score54

    Black Forest Labs releases FLUX.2 [klein] 4B Base on Hugging Face

    Black Forest Labs has published FLUX.2 [klein] 4B Base, a 4 billion parameter text-to-image model that also supports multi-reference editing. The model is undistilled, is released with open weights under Apache 2.0, and is described as fitting in about 13GB VRAM on cards such as the RTX 3090 or 4070, with reference code available in its GitHub repository and support in ComfyUI and Diffusers.

  2. Black Forest Labs · new models on Hugging FaceAI score46

    FLUX.2 [klein] 9B Base Released on Hugging Face as Undistilled Open-Weight Model

    Black Forest Labs has released FLUX.2 [klein] 9B Base, a 9 billion parameter undistilled rectified flow transformer with open weights for text-to-image generation and multi-reference editing. The model is intended for fine-tuning, LoRA training, and research, and fits in about 29GB VRAM on NVIDIA RTX 4090-class GPUs. A reference implementation is available on GitHub, and the model works with ComfyUI and Diffusers.

Dec 22, 2025

Dec 22, 2025Mon
  1. FunAudioLLM (Alibaba Tongyi) · new models on Hugging FaceAI score58

    Alibaba's FunAudioLLM releases Fun-Audio-Chat-8B for low-latency voice interaction

    FunAudioLLM has released Fun-Audio-Chat-8B, a roughly 8B-parameter large audio language model for natural, low-latency voice interaction, under Apache 2.0. It uses Dual-Resolution Speech Representations with a 5Hz frame rate, which the source says reduces GPU hours by nearly 50%, and it supports English and Chinese.

Dec 14, 2025

Dec 14, 2025Sun
  1. FunAudioLLM (Alibaba Tongyi) · new models on Hugging FaceAI score38

    Alibaba Releases Fun-ASR-MLT-Nano-2512, an 800M Multilingual Speech Recognition Model

    Alibaba's FunAudioLLM released Fun-ASR-MLT-Nano-2512, an 800M-parameter multilingual speech recognition checkpoint on Hugging Face that supports 31 languages, with emphasis on East and Southeast Asian languages. It is trained on hundreds of thousands of hours of speech and is available through the FunASR toolkit. The source's benchmark tables cover the Fun-ASR family rather than this checkpoint, so no checkpoint-specific accuracy figures are reported.

  2. FunAudioLLM (Alibaba Tongyi) · new models on Hugging FaceAI score36

    Fun-ASR-Nano-2512 Speech Recognition Model Released by Tongyi Lab on Hugging Face

    Tongyi Lab has released Fun-ASR-Nano-2512, an end-to-end speech recognition large model trained on tens of millions of hours of real speech, supporting low-latency real-time transcription across 31 languages. The model, which has 800M parameters, targets industry use such as education and finance and claims 93% accuracy in far-field, high-noise conditions. It is available on Hugging Face and works with the FunASR toolkit.

Dec 10, 2025

Dec 10, 2025Wed
  1. FunAudioLLM (Alibaba Tongyi) · new models on Hugging FaceAI score42

    Fun-CosyVoice3-0.5B-2512 Released as Open-Source Multilingual Text-to-Speech Model

    Alibaba's FunAudioLLM has released Fun-CosyVoice3-0.5B-2512, a 0.5B-parameter LLM-based text-to-speech model on Hugging Face, with an RL variant also published. The model supports zero-shot voice cloning across 9 languages and 18+ Chinese dialects and accents, with streaming output at latency as low as 150ms. On the source's test-en benchmark, it reports a 2.24% WER and 71.8% speaker similarity, and the RL version reports 1.68% WER.

Dec 4, 2025

Dec 4, 2025Thu
  1. ARC PrizeAI score62

    ARC Prize 2025 results point to refinement loops as the central AI reasoning trend

    ARC Prize reports that the top Kaggle entry reached 24% on the ARC-AGI-2 private dataset at $0.20 per task, and that all winning solutions and papers are open source. The top verified commercial model, Opus 4.5 (Thinking, 64k), scored 37.6% at $2.20 per task, while a Poetiq refinement on Gemini 3 Pro reached 54% at $30 per task. The author argues that refinement loops are the main driver of 2025 progress, and says ARC-AGI-3 is planned for early 2026.

    AIWhy it matters: The post links 2025 competition results to a broader argument about refinement loops, showing how benchmark outcomes are being read as evidence of AI reasoning progress.

Dec 2, 2025

Dec 2, 2025Tue
  1. Apple · new models on Hugging FaceAI score36

    Apple releases CLaRa-7B-E2E, an end-to-end RAG model with 16x and 128x compression

    Apple's CLaRa-7B-E2E is a fully end-to-end unified RAG model that jointly optimizes retrieval and generation, with 16x and 128x document compression. It is trained with end-to-end finetuning using differentiable top-k retrieval and a unified language-modeling objective. The model is available on Hugging Face with example end-to-end inference code.

Oct 30, 2025

Oct 30, 2025Thu
  1. Moonshot AI (Kimi) · new models on Hugging FaceAI score72

    Moonshot AI releases Kimi Linear 48B-A3B hybrid attention models on Hugging Face

    Moonshot AI has released Kimi-Linear-Base and Kimi-Linear-Instruct, both 48B total and 3B activated parameters with a 1M context length, on Hugging Face. The models use Kimi Delta Attention in a 3:1 hybrid ratio with global MLA, cutting KV cache by up to 75% and boosting decoding throughput by up to 6x at 1M tokens. The KDA kernel is open-sourced in FLA, and the checkpoints were trained on 5.7T tokens.

    AIWhy it matters: The model card gives concrete throughput and KV cache figures for a hybrid attention design, which helps readers weigh its long-context tradeoffs against full attention.

May 18, 2025

May 18, 2025Sun
  1. Cognition Blog (Devin, Windsurf)AI score22

    Cognition Revives Devin Open Source Initiative With $500 Credits for Projects

    Cognition is bringing back its Devin Open Source Initiative, offering $500 in Devin ACU credits to open-source GitHub projects with over 100 forks. Projects below that threshold will still be considered. Eligible projects must have an OSI-approved license and be actively maintained, and maintainers can apply through a linked form.

Jan 13, 2025

Jan 13, 2025Mon
  1. Cognition Blog (Devin, Windsurf)AI score38

    Crossmint Uses Devin to Scale Open-Source Development of GOAT SDK

    Crossmint said Devin became its top contributor to the open-source GOAT SDK during an initial trial, merging 8 pull requests versus 4 for the next contributor. Examples included a DEXScreener plugin built from a documentation URL and a harder Sui blockchain integration that needed three rounds of feedback and about an hour of human involvement. The company said its results depended on proper training, clear task context, and planned validation, not on treating Devin as superhuman.

Dec 11, 2024

Dec 11, 2024Wed
  1. Cognition Blog (Devin, Windsurf)AI score38

    Devin Open Source Initiative Gives Maintainers 500 Free ACUs for Repo Work

    Cognition is launching the Devin Open Source Initiative, giving selected open source maintainers 500 free ACUs on a Devin Teams plan as part of Devin's general availability launch. The post shows Devin contributing pull requests to projects including Anthropic's MCP Inspector, Dagger, and nanoGPT, with maintainers still reviewing the results. Devin's GitHub integration forwards PR comments and CI checks to help refine changes, though the company warns a human should still verify final quality.

Nov 30, 2024

Nov 30, 2024Sat
  1. Liquid AI BlogAI score60

    Liquid AI's STAR uses evolutionary search to synthesize tailored model architectures

    Liquid AI reports STAR, an evolutionary algorithm that synthesizes tailored neural network architectures from a numerical genome representation. The authors say it produced hundreds of designs that outperform Transformer and hybrid architectures in quality, with smaller caches and parameter counts, and can optimize for latency on target hardware. The full method is described in the arXiv technical report 2411.17800.

    AIWhy it matters: The post explains how evolutionary search over a new architecture design space produced designs beating Transformers and hybrids, giving a concrete method for quality versus latency and memory trade-offs.