タグ: R1

2025年5月下旬、中国のスタートアップ企業DeepSeekは、オープンソースのR1推論モデルの強化版であるR1-0528を静かにリリースした。このアップグレードでは、パラメータ数が6710億から6850億に増加し、軽量な蒸留バリアントが追加され、すべての重み、トレーニングレシピ、ドキュメントがMITライセンスの下でGitHubとHugging Faceで公開されている。ボンネットの下では、R1-0528は思考連鎖レイヤーを深くしてマルチステップロジックを改善し、学習後の微調整を適用して幻覚を抑制し、待ち時間を短縮している。統合面では、JSON出力、ネイティブ関数呼び出し、よりシンプルなシステム・プロンプトが導入され、数学、コーディング、ロジックのベンチマークにおいて、OpenAIのo3やGoogleのGemini 2.5 Proに匹敵するパフォーマンスを発揮する。

When DeepSeek dropped its R1 model in January 2025, it quietly detonated a bomb under the AI industry. Built without access to Nvidia’s best GPUs and on a shoestring budget compared to Silicon Valley giants, R1 proved that top-tier AI could come from unexpected places. Investors across the U.S. were caught flat-footed — nearly $600 billion was wiped off the tech markets as the realization set in: price-performance curves could collapse, not just bend. That was just the opening act. Now, DeepSeek is lining up two follow-ups — R2 and V4 — and if even half of the leaks are true, 2025 could be the year AI pricing falls off a cliff. The combination of massive model scale, brutally low prices, and zero […]

We’re thrilled to announce the release of Fello AI version 5.3.0, packed with exciting new features designed to empower your projects and boost productivity. This update brings two cutting-edge reasoning models—OpenAI o3-mini and DeepSeek R1 (hosted in the US)—as well as robust LaTeX support for beautifully rendered mathematical equations and formulas. OpenAI o3-mini & DeepSeek R1 The power of advanced reasoning is now at your fingertips with Fello AI. OpenAI o3-mini and DeepSeek R1, are engineered to handle complex, multi-step challenges—whether it’s solving technical puzzles, streamlining academic research, or optimizing data-driven decisions. What Are Reasoning Models? Most AI models are great at answering direct questions or summarizing simple content. But when things get more complex—like […]

Can an open-source AI really outsmart billion-dollar giants? Meet DeepSeek-R1, a groundbreaking model from China that’s turning heads worldwide. It recently scored an astonishing 97.3% on MATH-500, a math benchmark that leaves most competitors in the dust, and hit a 96.3% percentile on Codeforces, putting its programming skills on par with human experts. What’s even more surprising is that DeepSeek-R1 is fully open-source and costs a fraction of traditional models to train and deploy. With its reinforcement learning-based architecture and MIT license, it’s challenging the dominance of proprietary systems like OpenAI’s o1 and Google’s Gemini 2.0. This is more than just another AI release—it’s a revolution in accessibility, affordability, and […]