#模型发布

🚀 Meet MiniCPM5-2B, a 2B-parameter language model bringing high intelligence density to the edge, now open source! It ranks #1 among open-source models under 4B parameters on the @ArtificialAnlys Intelligence Index, with a score of 23. It also scores 20 on the Agentic Index, bringing an early form of general-purpose agent capability to the edge. Across 34 benchmarks, MiniCPM5-2B achieves an average score of 53.9, covering coding, math, long-context understanding, tool use, and agentic tasks. And this release goes beyond the model itself. We’re opening up the data, training recipes, and RL stack behind MiniCPM5-2B. 🤗 Hugging Face: http://huggingface.co/openbmb/MiniCPM5-2B 💻 GitHub: http://github.com/OpenBMB/MiniCPM Modelscope: http://modelscope.cn/models/OpenBMB/MiniCPM5-2B Web: https://openbmb.cn/

OpenBMB’s MiniCPM5 brings 128K context, coding, reasoning, and agentic tool use to on-device models.📜 Apache 2.0. 🤖 https://modelscope.ai/collections/OpenBMB/MiniCPM5 🏆 Both 1B&2B models reach open-source SOTA in their size classes. MiniCPM5-2B also beats the larger models in the official comparison. 🛠️ Switch between fast answers and deep reasoning. Native tool calling powers local assistants and coding agents. 🧠 RL + OPD merges 16 specialist models, including five Agent experts, into MiniCPM5-2B. 📂 Open training data includes 500K Agent SFT samples and 80K+ RL samples for code, math, knowledge, and long-context reasoning. ⚡ Standard Llama architecture works with major inference engines. Official cookbooks and Agent Skills cover deployment and fine-tuning across popular frameworks.

🚀 Sol-H3: @MiniMax_AI H3 Video Generation Faster Than Playback 🤩 Five seconds of world. 1.653 seconds to infer. We’re releasing Sol-H3, our fastest end-to-end MiniMax-H3 inference stack yet. On one 8× NVIDIA B300 Blackwell system, it generates five seconds of 1344×768 video with stereo audio in 1.653 seconds. Across 1×, 4×, and 8× B300, Sol-H3 reaches up to a 15.54× speedup versus Base H3. Compared with 50-step Base H3 Dense on the same 8× B300 system, the four-step Sol-H3 profile delivers: • 5s: 18.250s → 1.653s (11.04×) • 10s: 50.660s → 3.732s (13.57×) • 15s: 99.513s → 6.612s (15.05×) Sol-H3 also scales across GPU counts: • 4× B300: 2.918s / 6.993s / 12.542s for 5s / 10s / 15s (12.11–15.54×) • 1× B300: 13.745s / 37.813s / 52.260s for 5s / 10s / 15s (9.45–14.29×) All figures are medians of three measured runs after one warmup at 1344×768 and 24 FPS with stereo audio. Base H3 uses 50 scheduler points (49 DiT forwards); Sol-H3 uses four DiT forwards, so this is a full-profile comparison—not an attention-only runtime change. Sol-H3 uses Dense attention on 1× B300 and SOL with INT8 QKV / FP8 output transport on 4× / 8×. Timing includes text encoding, DiT denoising, and video/audio VAE decoding; model loading, compilation warmup, and final MP4 encoding are excluded. Sol-H3 brings Sol-Engine × Sol-Attn into one full-stack runtime: • dynamic sparse attention with no retraining • fused norm, RoPE, MLP, and sparse-attention setup • fused INT8 QKV / FP8 output communication across 8 GPUs • parallel, batched VAE decoding • precomputed AdaLN caching Inside the stack: • sparse-attention setup: 1.206 → 0.285 ms (−76.4%) • VAE decode: 7.55 → 0.602 s • ~24 GB memory freed per GPU Any MiniMax-H3 few-step LoRA can plug into the same engine, and the code is deployment-friendly under Apache 2.0. For us, the bigger milestone is crossing from “fast generation” into “faster than playback.” That opens the path toward continuous 24 FPS generation and truly interactive video systems. We’re excited to partner with @reactorworld to release Sol-H3 and make it available as an API day-0. Try it now on Reactor: https://www.reactor.inc/sandbox?model=fast-h3 🔗 https://nvlabs.github.io/Sana/Sol-Engine/Sol-H3/ Amazing team effort—full credits in the blog. @shanasaimoe @lawrence_cjs @yitongli165665 @haopengl33 @HaochengXiUCB @songhan_mit

Congrats @OpenBMB on launching MiniCPM5-2B! SGLang has day-0 serving support, with recipes for @NVIDIAAI RTX PRO 6000, RTX 5090, DGX Spark, and Hopper cards. We are seeing over 250 tok/s/user decode at bs=1 on coding tasks with DSpark enabled on a 5090. It's an ultra-efficient 2B with incredible intelligence density. As of today it leads Artificial Analysis's Intelligence Index among open models under 4B, and scores above models more than 4x its size. This is another step toward general-purpose models running 24x7 on device, even on phones. Run MiniCPM5-2B with SGLang 👇

🤝 Day-0 support for MiniCPM5-2B on stable vLLM. ⚡ Dense 2.6B model with 131K native context 🧠 Think / No-Think from the same checkpoint 🔧 Tool Calling support via vLLM’s minicpm5 parser Congrats @OpenBMB on the release, and thanks for keeping it on stock LlamaForCausalLM and opening the training data alongside the weights! 🙌 🔗 http://recipes.vllm.ai/openbmb/MiniCPM5-2B

OpenBMB's MiniCPM5-2B scores 15 on the Artificial Analysis Intelligence Index v4.2, the highest of any open weights model under 4B total parameters OpenBMB (@OpenBMB) is the open-source AI group behind the MiniCPM series of efficient small models. MiniCPM5-2B is a 2.6B parameter dense reasoning model with text input and output, released under Apache 2.0. Scoring 15 on the Intelligence Index, MiniCPM5-2B sits one point behind Ling 3.0 Tiny (16), which has ~3x the total parameters. Among open weights models under 4B total parameters, the next best score is Granite 4.2 3B (11). Key results: ➤ The highest Intelligence Index of any open weights model under 4B total parameters, setting a new Pareto-optimal point on Intelligence vs. Total Parameters: Its score of 15 is 4 points clear of Granite 4.2 3B (11). With 2.6B total parameters, it is 1 point ahead of Qwen3.5 4B (Reasoning, 14, estimated) with 44% fewer parameters, and level with Qwen3.5 9B (Reasoning, 15, estimated) at roughly 4x its size. As a dense model, its size advantage is in memory footprint rather than active-parameter compute. ➤ Strong agentic performance at this size: Its GDPval-AA v2 Elo of 831 leads <4B models, and on τ³-Banking it is joint-first with Ling 3.0 Tiny at 21%, compared to 8% for the next best model, Granite 4.2 8B. On AA-Briefcase, it placed second among the measured models in the comparison set with an Elo of 438, above Granite 4.2 8B (324) and just below Ling 3.0 Tiny (485). ➤ Knowledge, coding and long context are where it gives ground: MiniCPM5-2B places 7th in the set on Humanity's Last Exam (9%, behind Gemma 4 12B (Reasoning) at 16%), 8th on Terminal-Bench v2.1 (9%, behind Qwen3.5 9B (Reasoning) at 29%) and scores 0% on CritPt. On SciCode it is second of the five measured models at 26%, behind Granite 4.2 8B (31%). On AA-LCR v1.1 it scores 59%, 5th in the set, one point behind Ling 3.0 Tiny (60%). On GDP.pdf, our new professional document reasoning evaluation, it passes 1% of tasks outright, behind gpt-oss-20b (high) at 2%. ➤ Its AA-Omniscience score of -12 is earned by abstaining from answering rather than accuracy: MiniCPM5-2B attempts only 29% of AA-Omniscience questions, giving it a Non-Hallucination Rate of 78%. Its accuracy of 8% is a point below Ling 3.0 Tiny (9%) and half that of Qwen3.5 9B (Reasoning, 16%). Peers that attempt far more questions are penalized heavily, with Qwen3.5 9B (Reasoning) at -53 and gpt-oss-20b (high) at -63. ➤ It is token-efficient for a reasoning model: MiniCPM5-2B used 19k output tokens per Intelligence Index task, joint-lowest in the comparison model set with Granite 4.2 3B (19k). Ling 3.0 Tiny spends 56k, roughly 3x as many, for 1 more index point. Additional model details: ➤ Parameters: 2.6B (dense) ➤ Context window: 131k tokens ➤ Input modalities: Text only ➤ License: Apache 2.0

🚀 Hy4 preview is here. 770B, 49B active, 1M context. Built for productivity. Open source frontier. Consistent affordable price. Use it. Tell us what breaks. More on Hy blog：https://hy.tencent.ai/research/hy4-preview HuggingFace：https://huggingface.co/tencent/Hy4-preview Github：https://github.com/Tencent-Hunyuan/Hy4-preview

Ling-3.0-flash-Fin and FinFIRST are now open: a finance model built for real workflows, plus a benchmark that verifies the work. Developed by Ant Group with financial institutions and domain experts.🏦 🧠 MODEL 🤖 https://modelscope.ai/models/inclusionAI/Ling-3.0-flash-Fin ⚙️ 124B MoE, only 5.1B active per token, around 256K context. Financial training even lifts its AA Intelligence Index score from 38 to 41. 🛠️ Search sources, analyze filings, build valuations, update complex spreadsheets, and produce editable reports, all in private deployments. 🏆 82.45% on FinFIRST source verification, standing out among the open models evaluated. 📊 BENCHMARK 🤖 https://www.modelscope.ai/datasets/inclusionAI/FinFIRST 🔍 123 expert-authored tasks grade the complete research chain through 701 atomic criteria: not just answers, but sources, definitions, and reasoning.

GPT-6 Astra 现已面向 Work/Codex 中的所有 Pro、Enterprise 和 Business Premium 用户开放，并已在 API 中提供。 我们接下来将开始向 Plus 和 Business 用户推送。 感谢大家的耐心等待。