Cognition BlogSWE-2 is our most advanced coding model to date, achieving a score of 50.0% in the FrontierCode 1.1 Main evaluation, approaching the performance of Fable 5.1 with a 64% cost advantage, and pioneering the application of reinforcement learning to trillion-parameter scale.
Mistral AIMistral AI helps European energy operators migrate 40,000 lines of Fortran 77 code to C++, the physics-intensive reservoir simulator lacks a test suite and centralized documentation. The project verifies the numerical correctness of the migrated code by building a consistency test harness, and uses an AI agent to organize documentation, ultimately adopting a human-supervised agent workflow to achieve modular migration, significantly improving code quality.
OpenHands BlogWhen evaluating open-source large language models, it's necessary to distinguish between true open-source and open-weight models due to significant differences in weight disclosure, code, training data, and usage rights. Selection should focus on five points: license terms, the match between benchmark tests and real-world tasks, hardware requirements, contextual behavior, and ecosystem support.
Zhipu Z.ai · Hugging Face new modelsCompared to GLM-5.2, Zhipu GLM-5.3-BF16 improves coding ability by 50% on Z.ai Code Bench and achieves the best performance among open-source models on public benchmarks like Terminal Bench 3.0 and Agents' Last Exam; it shows top performance in vulnerability discovery tasks on CyberGym and scores more than double on the ExploitGym exploitation benchmark.
Liquid AI BlogLiquid AI experimentally verifies whether coding agents can autonomously solve production-level problems and open-sources tokenizer training tool toktoktok. The tool needs to handle trillions of tokens on a single machine. The experiment uses Claude Opus 4.5 and Codex with GPT-5.2 as agents, completing toy trainer development and passing unit tests within 30 minutes, but unable to process complete production datasets.
Liquid AI BlogLFM2.5-8B-A1B expands the vocabulary of the model tokenizer from 65K to 128K, reducing the number of tokens for Hindi, Vietnamese, and Thai texts by 2.4×, 2.6×, and 4.0× respectively, and improving device-side decoding speed by 2.2× to 3.7×. This expansion method extends the original BPE merging rules and fine-tunes the model in stages without retraining. The expanded 128K tokenizer has been released on Hugging Face, and the model itself has also been open-sourced.
Qoder Blog00Qoder releases #CaseFarewell to "Vibe Coding", introducing the practice evolution of team-level AI development through Harness governance and SDDAMAP framework. The framework supports model inference with an 8k context length, a 3.5x speedup, and a parameter size of 175B, achieving 95.2 on the MMLU benchmark test, currently provided as a closed-source API.
Qoder Blog1515CasePupu Supermarket and Qoder held a corporate-level AI programming workshop together, with over 150 developers participating in practical sessions with Qoder. The workshop focused on AI coding applications, demonstrating Qoder's integration capabilities in various scenarios such as IDE and CLI. Qoder provides enterprise pricing and Agent SDKs as development tools, supporting cloud agents and resource markets.
Qoder Blog00Qoder's five core features can improve efficiency, and six techniques can save points.
Meta AI BlogMuse Spark 1.1 is the latest multimodal reasoning model released by Meta Superintelligence Labs, significantly improving in tool and computer use, coding, and multimodal understanding, supporting million-token context windows. It is open to developers for public preview via the Meta Model API, surpassing Muse Spark and matching leading alternatives in Meta's internal coding evaluations.
Anthropic ResearchFeatured6767Why it matters: Based on approximately 400,000 Claude Code sessions, the division of labor between planning and execution was quantified, as well as how domain expertise affects task success rates.
Moonshot AI · Hugging Face new modelsKimi K2.7 Code is an encoding-focused intelligent agent built by Moonshot AI based on Kimi K2.6, showing significant improvements in real long-term encoding tasks, with about a 30% reduction in thinking token usage compared to Kimi K2.6.
Qoder Blog6565UU Paotui uses Qoder's Wiki, rules, and skills to transform AI from an 'outsider' into a full-lifecycle team member. The Qoder Agent SDK can be used for cloud and local integration and supports multiple development environments.
MiniMax · Hugging Face new modelsMiniMax-M3-MXFP8 is a native multimodal model with a 1M context window, approximately 428B parameters, and 23B activation parameters, achieving 9x pre-filling and 15x decoding speed improvement through MiniMax sparse attention mechanism (MSA). The model leads in long-horizon agent benchmarks, supporting three reasoning modes, and can be downloaded via Hugging Face and deployed using SGLang, vLLM, or Transformers frameworks.
Anthropic Research81% of social science researchers have tried using AI chatbots for research, but only 20% have adopted coding proxy tools like Claude Code. Researchers who adopt coding proxies are more likely to come from top universities or male-named researchers, and they publish more papers and work proposals, but this may also reflect prior differences among early adopters. AI coding proxies are the first to make core research tasks automatable, which may reshape social science research but could also exacerbate research resource inequality.
Prime Intellect Blog6565Prime Intellect takes FrontierSWE to the Environments Hub, an ultra-long temporal open-source coding benchmark developed by Proximal, which includes challenges such as optimizing compilers or training SOTA protein prediction models, with an average task duration of 11 hours and almost unsolvable.
Cognition BlogThe SWE-Check model achieves performance comparable to Opus 4.6 in internal evaluations (F1 score increased from 0.09 to 0) and shows significant improvement in external evaluations (F1 score increased from 0.49 to 0.29), with 10 times faster runtime and lower inference costs. The model, developed through integrating native production environment integration, reward linearization techniques, and multi-stage fine-tuning, is available for preview in Windsurf Next and will be officially released soon.
Qoder Blog6565A photographer used Qoder to write 150,000 lines of code and independently delivered 3 commercial products in 5 months.
Qoder Blog6565Qoder NEXT proactively identifies developers' coding intent, providing smarter code editing suggestions, significantly improving development efficiency. Its Agent SDK supports cloud agent integration, with a context length of up to 128k tokens, inference speed 3.5 times faster than baseline models, priced at $3/month/M tokens, currently provided as a closed-source API.
Anthropic ResearchFeatured7878Harvard physics professor Matthew Schwartz guides Claude Opus 4.5 to perform NLL re-sum calculation of Sudakov shoulder in C-parameter, producing a high-energy theoretical physics thesis containing a new factorization theorem within two weeks, while similar work usually takes a year.
Why it matters: Harvard physics professor uses two weeks, 110 drafts, and 36M tokens to let Claude complete a high-energy physics thesis, listing point by point what the model can and cannot do.
Qoder Blog2222Qoder has launched the Expert Mode, transforming AI from a 'programmer' to an 'engineering expert team', achieving delivery capabilities beyond coding. This mode supports context lengths of 8k tokens, with a 3.5x speed increase in reasoning, priced at $3.5 per million tokens, and integrated via API and SDK.
Andrej Karpathy released an art project named microgpt, a single-file containing only 200 lines of pure Python code, which can train and infer GPT models without dependencies. The project implements all algorithmic content from dataset processing, tokenizer, automatic differentiation engine to GPT-2 style neural network architecture, Adam optimizer, training and inference loop. The code can be obtained from GitHub gist and karpathy.art, and has been released as a triptych.
Qoder Blog00The Qoder NEXT model upgrades editing capabilities from code completion to prediction, achieving more precise intelligent editing through ActionRL preference alignment technology. The model's context length reaches 128k, with a parameter size of 175B, demonstrating excellent performance in the SWE-bench reasoning benchmark, a 3.5x speed increase, and supporting open-source API interfaces.
Qoder Blog00Qoder NEXT achieves millisecond-level code completion response through optimization, and its Agent SDK supports multiple integration methods, suitable for enterprise applications.
Anthropic ResearchFeatured6767Anthropic's research team developed an agent based on Claude Code custom commands, which infers the properties that code should satisfy by reading type annotations, docstrings, and comments, and then uses Hypothesis to write property tests, discovering hundreds of potential bugs in over 100 popular Python packages such as NumPy, SciPy, and Pandas.
Why it matters: Anthropic uses Claude agents to automatically write property tests to find bugs in packages like NumPy, SciPy, Pandas, and provides the efficiency after manual review and merged fixes
Berkeley AI ResearchBerkeley AI research proposes an information-driven imaging system design framework, which quantifies the object's discrimination capability through mutual information measurement. It has been validated in four fields: color photography, radio astronomy, lensless imaging, and microscopy that information estimation can predict decoder performance. The optimized mutual information design scheme achieves end-to-end optimal results while reducing memory and computational load.
Qoder Blog00Qoder introduces a hybrid method of code retrieval using a technical codebase perception code search, achieving AI coding through general embedding into real-time graph-driven codebase understanding. The model parameter size is 175B, achieving 95.2% accuracy on the SWE-bench validation set, a 3.5x speed increase, and a context window supporting 8k tokens, currently provided in a closed-source API format.
Qoder Blog6565Qoder introduces the ProductQuest mode, delegating tasks to AI-assisted coding. This mode supports a context length of 128k tokens and speeds up the process 3.5 times.
Cognition BlogSWE-1.5 is a boundary-scale model with hundreds of billions of parameters, achieving near SOTA performance in coding, and its service speed has been increased to up to 950 tok/s through collaboration with Cerebras, 6 times faster than Haiku 4.5 and 13 times faster than Sonnet 4.5. SWE-1.5 is now available on Windsurf.
Anthropic ResearchFeatured6262Anthropic releases Claude Code best practices covering CLAUDE.md configuration, tool permission management, MCP access, common workflows, and multi-Claude parallelism. The article suggests exploring and planning with sub-agents before writing code and provides specific processes such as test-driven development, screenshot iteration, and headless mode automation. This article has been moved to the official documentation, and the original blog is no longer maintained.
Why it matters: Anthropic's official system has organized the configuration, workflows, and multi-instance parallel methods of Claude Code, which can be used as a reference for setting up intelligent agent coding processes.
Dify Blog6565Dify's low-code platform combined with DeepSeek R1 can quickly build enterprise-level multilingual document translation tools. Through a visual workflow, it achieves document parsing, stylized translation, and batch processing, enhancing cross-language collaboration efficiency without complex programming.
Dify Blog6565Dify and Windsurf created a Twitter MBTI analyzer, gaining 200,000 visits in the first three days of launch. The project uses Dify to build the backend, analyzes user profiles with Twitter Scraper to generate MBTI reports, and the frontend is developed by Windsurf's Claude-3.5-sonnet assistant. API communication uses HTTP requests, supporting an 8k context window.
Liquid AI BlogThe STAR algorithm synthesizes customized architectures through evolutionary algorithms, with the generated architectures being superior in quality to strong Transformers and hybrid architectures, using fewer parameters and cache. STAR can customize architectures based on tasks, metrics, and hardware, employing the theory of linear input variation systems (LIVs) and hierarchical numerical sequences (STAR genomes) for evolutionary optimization, supporting both static and dynamic objectives.
Dify BlogWhen developers use OpenAI's GPT API, the AI may not produce the expected output or may 'be uncooperative'. To address these challenges and improve GPT performance, it is necessary to follow three key steps: guide model behavior through prompt engineering, use single-point or few-shot learning techniques, and leverage embeddings to extend model context, and can adapt the model to specific domains through fine-tuning.
