Ricursive 发布 Ripple-65B 预览版
Ricursive Intelligence 刚刚发布了 Ripple-65B Preview。Ripple-65B Preview 是 Ricursive 推出的 65B 稠密开放权重模型,从零开始预训练,面向智能体式编程。它是 Ricursive 的 AI 芯片设计平台背后专有模型的通用版本。其核心宣称是:在 Terminal-Bench 2.1 上取得 80.52 分,是美国开放权重模型中自报成绩最高的。在该测试中被它击败的每一个 MoE 对手,总参数量都在 501B 到 975B 之间。
此次发布包含 3 个 checkpoint:一个指令微调的 BF16 模型、一个 NVFP4 版本,以及一个 BF16 基础模型。全部以 Apache 2.0 许可发布。预训练使用的是纯文本语料,以英文和代码为主,序列长度为 8K。
Ricursive 团队表示,全部研发成本为 3200 万美元。该数字涵盖 ladder 实验、数据配比、评测、合成数据以及最终训练轮次。Ricursive 团队估计,这是此前开放权重智能体式编程领先模型成本的十分之一。
Ricursive 选择稠密 Transformer 而非混合专家(MoE),以降低基础设施复杂度。训练在简单分片下保持了 40% 的 MFU,未使用自定义 kernel,也未做量化。
该模型又瘦又深:d_model 为 5120,共 80 层。它从 Olmo 3 类似 Llama 的布局转向类似 Gemma 3N 的设计,以提升稳定性。关键选择如下:
Ricursive 团队从 Olmo 3 的 ladder 出发,通过 ladder 实验构建了这套配方。他们舍弃了在 4B 规模上无法保持有效性的改动,例如 2-stream mHC。技术文章称,在 4B 规模上取得了 16 倍的代码计算效率提升。
后训练依次进行了 midtraining、长上下文、SFT、长上下文 SFT 和混合 RL。长上下文阶段以 2M 序列长度训练。Ricursive 没有购买任何后训练数据,而是自建合成数据流水线。
该强化学习方案结合了截断重要性采样、分数中心化和线性长度惩罚。Ripple 在 Terminal-Bench 2.1 上取得 80.5 分,用了 140 万次 rollout。Ricursive 指出,这比 Beam 少了约 70 倍,后者报告的 rollout 超过 1 亿次。
Ricursive 团队在内部用 OpenCode 补跑了缺失的评测,并用 Terminus 2 完成 SWE-bench Pro v2.0 Hard 和 DeepSWE 1.1。以下分数均为 Ricursive 提供的数据。
Ricursive 团队还用官方 Terminus 2 测试框架在 128K 下重新跑了 Terminal-Bench 2.1。多数对手在该设置下成绩下滑;Beam 降至 70.8。Muse Spark Open 升至 78.7,仍低于 Ripple 的 80.5。
译文由英文原文机器翻译,术语按行业惯例统一,可在文末展开原文核对。
查看英文原文(Ricursive AI Releases Ripple-65B Preview: A 65B Dense Open-Weight Model Scoring 80.5 on Terminal-Bench 2.1)
Ricursive Intelligence just released Ripple-65B Preview. Ripple-65B Preview is a 65B dense open-weight model from Ricursive, pretrained from scratch for agentic coding. It is a general-purpose variant of the proprietary model behind Ricursive’s AI chip design platform. The main claim: 80.52 on Terminal-Bench 2.1, the top self-reported score among US open-weight models. Every MoE rival it beats on that test carries 501B to 975B total parameters.
TL;DR
- Size: 65B dense (65B active), pretrained on 12T tokens, up to 1M context.
- Runs on: BF16 on datacenter GPUs (250 tok/s on 2x B200). Experimental NVFP4 build targets NVIDIA DGX Spark.
- Performance: Frontier-level agentic coding for its size; trails larger MoE models on knowledge and science coding.
- Best: Terminal-Bench 2.1, 80.5 (self-reported).
- Worst: SciCode, 39.4, lowest of the 5 models in Ricursive’s own table.
What is Ripple-65B Preview?
The release includes 3 checkpoints: an instruction-tuned BF16 model, an NVFP4 version, and a BF16 base model. All ship under Apache 2.0. Pretraining used a text-only corpus, mostly English and code, at 8K sequence length.
Ricursive team states that the full R&D cost was $32M. That figure covers ladder experiments, data mixing, evals, synthetic data and the final runs. Ricursive team estimates that it is one-tenth the cost of the prior open-weight agentic coding leader.
How does the architecture work?
Ricursive chose a dense Transformer over Mixture-of-Experts to reduce infrastructure complexity. Training held 40% MFU with simple sharding, no custom kernels and no quantization.
The model is thin and deep: a d_model of 5120 across 80 layers. It moves from Olmo 3’s Llama-like layout to a Gemma 3N-like design for stability. Key choices:
- Overprovisioned attention heads for reasoning, offset by more sliding-window layers and smaller windows.
- Gated attention, global NoPE and value residuals.
- FFN activation clipping, a technique also used in gpt-oss.
Ricursive team built the recipe through laddering experiments starting from Olmo 3’s ladder. It rejected changes that failed to hold at 4B scale, such as 2-stream mHC. The technical post reports a 16x code compute-efficiency gain at 4B scale.
How was it post-trained?
Post-training ran midtraining, long context, SFT, long-context SFT and mixed RL. Long-context stages trained at 2M sequence length. Ricursive bought no post-training data and built its own synthetic pipelines.
The RL recipe combines truncated importance sampling, score centering and a linear length penalty. Ripple hit 80.5 on Terminal-Bench 2.1 with 1.4M rollouts. Ricursive puts that at roughly 70x fewer than Beam, which reported over 100M rollouts.
How does Ripple-65B compare on benchmarks?
Ricursive team ran missing evals in-house using OpenCode, and Terminus 2 for SWE-bench Pro v2.0 Hard and DeepSWE 1.1. Scores below are Ricursive’s presented data.
- Terminal-Bench 2.1: 80.5, vs Beam 80.1, Muse Spark Open 76.4, Inkling 63.8, Nemotron 3 Ultra 56.4.
- SWE-bench Pro v2.0 Hard: 79.1, the top score in Ricursive’s table.
- SWE-bench Verified: 76.5, below Beam and Inkling.
- Math: 98.3 on AIME 2025 and 90.5 on HMMT Feb 2026.
- Weak spots: SciCode 39.4, DeepSWE 1.1 44.3, MCP Atlas 71.67, GPQA Diamond 85.2.
Ricursive team also reran Terminal-Bench 2.1 with the official Terminus 2 harness at 128K. Most rivals dropped there; Beam fell to 70.8. Muse Spark Open rose to 78.7, still below Ripple’s 80.5.
Comparison table
| Feature | Ripple-65B Preview | Reflection Beam | Meta Muse Spark Open | Thinking Machines Inkling | NVIDIA Nemotron 3 Ultra |
|---|---|---|---|---|---|
| Architecture | Dense | MoE | MoE | MoE | Hybrid Mamba-2 LatentMoE |
| Total params | 65B | 501B | 816B | 975B | 550B |
| Active params | 65B | 23B | 42B | 41B | 55B |
| Pretraining tokens | 12T | 23.8T | Not disclosed | 45T (per Ricursive table) | ~20T |
| Context | Up to 1M | 1M (midtraining) | 1M | 1M | Up to 1M |
| License | Apache 2.0 | Apache 2.0 | Muse Model License | Apache 2.0 | OpenMDW-1.1 |
| Input modality | Text | Text | Text, image, video, audio | Text, image, audio | Text |
| Self-hosting floor | DGX Spark class (NVFP4) | Not disclosed | 700GB VRAM (NVFP4) | 600GB VRAM (NVFP4) | 8x B200 or 16x H100 |
| Terminal-Bench 2.1 | 80.5 | 80.1 | 76.4 | 63.8 | 56.4 |
| SWE-bench Verified | 76.5 | 80.9 | Not disclosed by Meta | 77.6 | 70.7 |
| GPQA Diamond | 85.2 | Not disclosed in blog | 89.0 | 87.2 | 87.0 |
Ripple scores from the Ricursive model card. Competitor scores from each vendor’s own source. All self-reported.
Where can developers run it?
Weights are on Hugging Face. Ricursive lists Together AI, GMI Cloud and OpenRouter at launch, with Fireworks later. Together AI’s page currently marks the model as coming soon.
On datacenter GPUs, the model reaches 250 tok/s on 2x B200 at batch 8, before quantization. The experimental NVFP4 checkpoint quantizes MLP layers and keeps attention in BF16. Ricursive team positions it for local devices like NVIDIA DGX Spark. A DFlash2 variant is also listed.
Key Takeaways
- 65B dense, 12T tokens, 1M context, Apache 2.0.
- 80.5 on Terminal-Bench 2.1, top among US open weights.
- $32M full R&D, 1.4M RL rollouts.
- NVFP4 build targets DGX Spark-class local hardware.
- Trails larger MoE rivals on SciCode, DeepSWE and MCP Atlas.
Check out the the launch blog, the model card and the NVFP4 checkpoint. Also, feel free to follow us on Twitter and don’t forget to join our 150k+ML SubReddit and Subscribe to our Newsletter. Wait! are you on telegram? now you can join us on telegram as well.
Need to partner with us for promoting your GitHub Repo OR Hugging Face Page OR Product Release OR Webinar etc.? Connect with us
The post Ricursive AI Releases Ripple-65B Preview: A 65B Dense Open-Weight Model Scoring 80.5 on Terminal-Bench 2.1 appeared first on MarkTechPost.
正文保留英文原文,标题与摘要由机器翻译。
原文发布于 MarkTechPost(Ricursive AI Releases Ripple-65B Preview: A 65B Dense Open-Weight Model Scoring 80.5 on Terminal-Bench 2.1),标题、摘要与正文由机器翻译,术语按行业惯例统一,未经人工逐句校订。