Head to head · Open-weight models you run yourself · October 2026 research run
Kronos vs Ripple-65B Preview
Kronos scores 55.2 (C) on agent readiness against Ripple-65B Preview's 53.3 (D), and leads in 2 of 4 scored categories. Ripple-65B Preview leads on risk and agent interface. Both do open-weight models you run yourself.
Best open-weight models you download and run yourself for AI agents · All 3674 open weights comparisons
Which one, for what
Kronos C
Best for Agents that need a quick local forecast or synthetic sample of a candlestick series on modest hardware, for research or as one input among others.
Ahead on
- Reach, 55 against 45
- Documentation and upkeep, 61 against 49
Watch for
Runs only through the publisher's own PyTorch code from GitHub, in no named runtime and with no quantised build
Best for Teams with Blackwell GPUs or a DGX Spark who want an Apache 2.0 coding model to try for terminal and agentic coding work, and can run the publisher's code to serve it.
Ahead on
- Risk, 55 against 50
- Agent interface, 75 against 60
Watch for
No released runtime runs the architecture, so running it means Ricursive's Python in SGLang 0.5.19 or Transformers' remote code
Why the edge
Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.
Agent interface Ripple-65B Preview +15
Kronos 60
Read for a forecasting model, the tool-calling, trained tool calling and structured-output lines stand for what it takes and returns. The card documents the DataFrame and timestamps it takes and the DataFrame it returns, and the publisher's KronosPredictor builds and parses them (25). It was trained on six K-line fields with calendar fields, one series per forecast (10). Output is a typed DataFrame indexed by the requested timestamps, the mean of the sampled paths (15). The input format is given only as code (10). Reasoning control and context score 0, with no reasoning and a 512-bar context (2,048 for Kronos-mini).
Ripple-65B Preview 75
Chat template 20, chat_template.jinja in the repository. Tool-calling format 25, Ricursive's server README says how tools are declared and calls come back, with the ripple parser in SGLang 0.5.19. Trained tool calling 20, one or more calls a turn and agentic-coding RL. Structured output 0, not documented. Reasoning control 0, the template has no thinking switch or effort levels. Context 10, long-context training at 2M-token sequences for a stated 1M.
Documentation and upkeep Kronos +12
Kronos 61
Model card 11, intended use, architecture, parameters and context are there, limitations, languages and a cut-off aren't. Training data 20, the paper lists every exchange with asset and observation counts, the filtering and a June 2024 end. Last change 5, the reference code on 13 April 2026 (still the newest commit on 4 November, 205 days) and the Hub repositories on 9 September 2025. Responsiveness 20, the publisher answered 10 of 18 Hugging Face discussions (recounted by script on 4 November), though none since 24 November 2025, and the seven opened since April 2026 have no reply from it. Version policy 5, a first release. File health 0, the cards point to example data the repository has deleted and to licence and requirements links on a branch that doesn't exist.
Ripple-65B Preview 49
Model card 14 of 20, no limitations and no knowledge cut-off. Training data 5, a general description with token counts and no cut-off. Last change 20, released today. Responsiveness 0, no discussions yet, read as none answered. Version policy 5, a first release, with the NVFP4 checkpoint replaced in place on release day under a commit note. File health 5, tokenizer_config.json carries a second, tool-less chat template and a 131,072-token limit against the card's 1M.
Reach Kronos +10
Kronos 55
Memory 30, the publisher's Kronos-mini with its tokeniser is 0.03 GB and even Kronos-base is 0.43 GB in FP32. Runtimes 0, it runs only through the publisher's own PyTorch package from GitHub, and none of the six named runtimes supports it. Builds 0, FP32 safetensors only. Terms 15, MIT with nothing added. Download 10, nothing is gated. Hosted 0, no inference provider or publisher API serves it (provider mapping still empty on 4 November).
Ripple-65B Preview 45
Memory 10 of 30, the smallest 4-bit build is Ricursive's NVFP4 at 57.2 GB, 68.7 GB with a fifth added, which is over 64 GB and fits our reference machine. Runtimes 0, no released runtime runs the ricursive architecture with its own code; Ricursive's SGLang plugin and Transformers' remote code don't count. Builds 10, one quantised precision (NVFP4) from Ricursive. Terms 15, Apache 2.0 with nothing added. Download 10, not gated. Hosted 0, Ricursive serves no API, Hugging Face maps no provider, Together AI's page says coming soon, and GMI Cloud's and OpenRouter's pages weren't read.
Risk Ripple-65B Preview +5
Kronos 50
File safety 15, safetensors only, but the only documented loading path is the publisher's Python package in place of a runtime. Provenance 15, the files are in the NeoQuasar account the publisher's README links, with SHA-256 hashes. Safety evaluation 0, none published. Agent guidance 5, the README's general cautions that the fine-tuning pipeline isn't a production trading system. Licence risk 15, plain MIT; the broken licence link is charged to file health. Security programme 0, no reporting route found.
Ripple-65B Preview 55
File safety 15, safetensors only, less 10 because the documented way to run it loads Ricursive's Python. Provenance 15, the files are in the organisation ricursive.com links to, with a checksum on every file. Safety evaluation 10, Ricursive's own FORTRESS and StrongREJECT results. Agent guidance 0, nothing on prompt injection, refusals or safeguards. Licence risk 15, Apache 2.0. Security programme 0, no way to report a flaw.
Score by category
| Category | Weight this run | Kronos | Ripple-65B Preview | Edge |
|---|---|---|---|---|
| Capability | 35%pending | pending | pending | not scored in this run |
| Reach | 25%25 | 55 | 45 | Kronos +10 |
| Risk | 20%20 | 50 | 55 | Ripple-65B Preview +5 |
| Agent interface | 10%10 | 60 | 75 | Ripple-65B Preview +15 |
| Documentation and upkeep | 10%10 | 61 | 49 | Kronos +12 |
| Negative events | ≤15 | 0 | 0 | |
| Total | C 55.2/100 | D 53.3/100 |
Facts side by side
| Fact | Kronos | Ripple-65B Preview |
|---|---|---|
| Kind | Open-weight model | Open-weight model |
| Vendor | Yu Shi (shiyu-coder, NeoQuasar), Tsinghua University | Ricursive Intelligence |
| Hosted endpoint | no (local only) | no (local only) |
| Transports | ||
| Auth | None | None |
| Pricing | Free | Free |
| x402 | no | no |
| Licence | MIT | Apache-2.0 |
| Read-only variant documented | no | no |
| llms.txt | no | no |
| Last release | 2026-04-13 | 2026-11-04 |
| Terms last updated | no document linked | not read yet |
| Privacy policy last updated | no document linked | no document linked |
| Popularity | 43k stars | none |
Verdicts
Kronos
Forecasts the next price bars from a history of up to 512 candlestick bars. Every size fits in under 0.5 GB and runs on a CPU, but only through the publisher's own Python package. Nobody has benchmarked it independently, its quickstart data was deleted from the repository, and the publisher hasn't replied since November 2025.
Ripple-65B Preview
A dense 65B coding model from a new lab, pretrained from scratch and released today under Apache 2.0, that Ricursive says leads U.S. open-weight models on Terminal-Bench 2.1. Nobody has checked that yet, and the custom architecture runs only through Ricursive's own SGLang plugin on Blackwell GPUs, with a 57 GB NVFP4 build as the smallest.
Before you call either
Kronos
- Clone https://github.com/shiyu-coder/Kronos and
pip install -r requirements.txt, then loadKronos.from_pretrained("NeoQuasar/Kronos-small")andKronosTokenizer.from_pretrained("NeoQuasar/Kronos-Tokenizer-base")and wrap them in KronosPredictor - Pair Kronos-mini with Kronos-Tokenizer-2k and Kronos-small or Kronos-base with Kronos-Tokenizer-base, the cards' pairing
- The examples read examples/data/XSHG_5min_600977.csv, which no longer exists; point them at tests/data/regression_input.csv, 2,500 five-minute bars
- Keep the lookback at or under max_context, 512 bars, and pass the future timestamps you want as y_timestamp, since the calendar fields come from them
- Raise sample_count for a steadier forecast, but expect only the mean back; call the lower-level auto_regressive_inference yourself if you need the paths
- Treat forecasts as model output to check, not signals; the README says its own backtest pipeline isn't a production trading system
Ripple-65B Preview
- Run Ricursive's launcher from inference/server with
uv run --script serve.pyon two B200 GPUs; it listens on 127.0.0.1:30000 with a 16K context by default - The launcher turns on
--reasoning-parser rippleand--tool-call-parser ripple; without them tool calls come back as raw text inside tool_call tags - For
tool_choice: "required"setstrict: trueandadditionalProperties: falseon each function, which the README says stopped extra argument fields - Leave room in
max_tokensfor reasoning as well as the answer, since thinking is always on - Use chat_template.jinja, not the template inside tokenizer_config.json, which has no tool support
Questions
Which is better for AI agents, Kronos or Ripple-65B Preview?
Kronos scores 55.2 (C) on agent readiness against Ripple-65B Preview's 53.3 (D), and leads in 2 of 4 scored categories. Ripple-65B Preview leads on risk and agent interface.
Other comparisons with Kronos or Ripple-65B Preview
- Agents-A1-4B vs Kronos
- Agents-A1-4B vs Ripple-65B Preview
- Apertus v1.5 8B vs Kronos
- Apertus v1.5 8B vs Ripple-65B Preview
- Apodex-1.1-mini vs Kronos
- Apodex-1.1-mini vs Ripple-65B Preview
- BigBang-v1 vs Kronos
- BigBang-v1 vs Ripple-65B Preview
- Carbon-X vs Kronos
- Carbon-X vs Ripple-65B Preview
- Cosmos-Reason2-2B vs Kronos
- Cosmos-Reason2-2B vs Ripple-65B Preview
- Darwin-36B-Opus vs Kronos
- Darwin-36B-Opus vs Ripple-65B Preview
- DeepSeek-V3.2 vs Kronos
- DeepSeek-V3.2 vs Ripple-65B Preview
- DeepSeek V4.1 Flash vs Kronos
- DeepSeek V4.1 Flash vs Ripple-65B Preview
- DeepSeek-V4-Flash-0731 vs Kronos
- DeepSeek-V4-Flash-0731 vs Ripple-65B Preview
- DiffusionGemma 26B A4B IT vs Kronos
- DiffusionGemma 26B A4B IT vs Ripple-65B Preview
- Gemma 4 12B IT vs Kronos
- Gemma 4 12B IT vs Ripple-65B Preview
- Gemma 4 26B A4B IT vs Kronos
- Gemma 4 26B A4B IT vs Ripple-65B Preview
- Gemma 4 31B IT vs Kronos
- Gemma 4 31B IT vs Ripple-65B Preview
- Gemma 4 E2B IT vs Kronos
- Gemma 4 E2B IT vs Ripple-65B Preview
- Gemma 4 E4B IT vs Kronos
- Gemma 4 E4B IT vs Ripple-65B Preview
- GLM-4.5-Air vs Kronos
- GLM-4.5-Air vs Ripple-65B Preview
- GLM-4.7-Flash vs Kronos
- GLM-4.7-Flash vs Ripple-65B Preview
- GLM-5.3-Flash vs Kronos
- GLM-5.3-Flash vs Ripple-65B Preview
- GLM-5.3 vs Kronos
- GLM-5.3 vs Ripple-65B Preview
- gpt-oss-120b vs Kronos
- gpt-oss-120b vs Ripple-65B Preview
- gpt-oss-20b vs Kronos
- gpt-oss-20b vs Ripple-65B Preview
- Granite 4.2 3B vs Kronos
- Granite 4.2 3B vs Ripple-65B Preview
- Hy3 vs Kronos
- Hy3 vs Ripple-65B Preview
- Inkling-Small vs Kronos
- Inkling-Small vs Ripple-65B Preview
- Inkling vs Kronos
- Inkling vs Ripple-65B Preview
- InternVL3.5-GPT-OSS-20B-A4B-Preview vs Kronos
- InternVL3.5-GPT-OSS-20B-A4B-Preview vs Ripple-65B Preview
- Kimi K3 vs Kronos
- Kimi K3 vs Ripple-65B Preview
- Kolibri-1 vs Kronos
- Kolibri-1 vs Ripple-65B Preview
- Kronos vs Laguna M.1
- Kronos vs Laguna S 2.1
- Kronos vs Laguna XS 2.1
- Kronos vs LBS-27B
- Kronos vs LFM2.5-2.6B
- Kronos vs Ling-3.0-tiny
- Kronos vs MiMo-V2.6 Distill Qwen 9B
- Kronos vs MiniCPM-o 4.5
- Kronos vs MiniCPM5-2B
- Kronos vs MiniMax-M2.5
- Kronos vs MiniMax-M2.7
- Kronos vs MiniMax-M3.1-Flash
- Kronos vs Mistral Large 4
- Kronos vs Muse Glimmer 30B
- Kronos vs Nanbeige4.3-33B
- Kronos vs NVIDIA Nemotron 3.5 Bolt E4B 4GB
- Kronos vs Nemotron 3.5 Lightning 30B A3B
- Kronos vs Nemotron 3 Nano 30B A3B
- Kronos vs Nemotron 3 Nano 4B
- Kronos vs Nemotron 3 Nano Omni 30B A3B Reasoning
- Kronos vs Nemotron 3 Super 120B A12B
- Kronos vs Nemotron 3 Ultra 550B A55B
- Kronos vs occamy-1.0
- Kronos vs Olmo 3 7B Think
- Kronos vs Ornith-1.5-35B-A3B
- Kronos vs Ornith-1.5-397B
- Kronos vs Ornith-1.5-9B
- Kronos vs Qwen3-30B-A3B-Instruct-2507
- Kronos vs Qwen3-4B-Instruct-2507
- Kronos vs Qwen3.5-0.8B
- Kronos vs Qwen3.5-122B-A10B
- Kronos vs Qwen3.5-2B
- Kronos vs Qwen3.5-397B-A17B
- Kronos vs Qwen3.5-4B
- Kronos vs Qwen3.5-9B
- Kronos vs Qwen3.6-27B
- Kronos vs Qwen3.6-35B-A3B
- Kronos vs Qwen3.8-27B
- Kronos vs Qwen3.8-Flash-Next
- Kronos vs Qwen3-Coder-30B-A3B-Instruct
- Kronos vs Qwen3-Coder-Next
- Kronos vs Qwen3-Omni-30B-A3B-Instruct
- Kronos vs Qwen3-VL-2B-Instruct
- Kronos vs Qwen3-VL-30B-A3B-Instruct
- Kronos vs Qwen3-VL-32B-Instruct
- Kronos vs Qwen3-VL-4B-Instruct
- Kronos vs Qwen3-VL-8B-Instruct
- Kronos vs Qwopus3.8-27B-Flash-V2
- Kronos vs Recursive Flint
- Kronos vs SmolLM3-3B
- Kronos vs Spark-X2.6-4B
- Kronos vs Swift 1.5 Qwen3.8-27B
- Kronos vs Swift 1.5 Qwen3.8-Flash-Next
- Kronos vs ThinkingCap (Qwen3.8 27B)
- Kronos vs Baidu Unlimited-OCR-2
- Laguna M.1 vs Ripple-65B Preview
- Laguna S 2.1 vs Ripple-65B Preview
- Laguna XS 2.1 vs Ripple-65B Preview
- LBS-27B vs Ripple-65B Preview
- LFM2.5-2.6B vs Ripple-65B Preview
- Ling-3.0-tiny vs Ripple-65B Preview
- MiMo-V2.6 Distill Qwen 9B vs Ripple-65B Preview
- MiniCPM-o 4.5 vs Ripple-65B Preview
- MiniCPM5-2B vs Ripple-65B Preview
- MiniMax-M2.5 vs Ripple-65B Preview
- MiniMax-M2.7 vs Ripple-65B Preview
- MiniMax-M3.1-Flash vs Ripple-65B Preview
- Mistral Large 4 vs Ripple-65B Preview
- Muse Glimmer 30B vs Ripple-65B Preview
- Nanbeige4.3-33B vs Ripple-65B Preview
- NVIDIA Nemotron 3.5 Bolt E4B 4GB vs Ripple-65B Preview
- Nemotron 3.5 Lightning 30B A3B vs Ripple-65B Preview
- Nemotron 3 Nano 30B A3B vs Ripple-65B Preview
- Nemotron 3 Nano 4B vs Ripple-65B Preview
- Nemotron 3 Nano Omni 30B A3B Reasoning vs Ripple-65B Preview
- Nemotron 3 Super 120B A12B vs Ripple-65B Preview
- Nemotron 3 Ultra 550B A55B vs Ripple-65B Preview
- occamy-1.0 vs Ripple-65B Preview
- Olmo 3 7B Think vs Ripple-65B Preview
- Ornith-1.5-35B-A3B vs Ripple-65B Preview
- Ornith-1.5-397B vs Ripple-65B Preview
- Ornith-1.5-9B vs Ripple-65B Preview
- Qwen3-30B-A3B-Instruct-2507 vs Ripple-65B Preview
- Qwen3-4B-Instruct-2507 vs Ripple-65B Preview
- Qwen3.5-0.8B vs Ripple-65B Preview
- Qwen3.5-122B-A10B vs Ripple-65B Preview
- Qwen3.5-2B vs Ripple-65B Preview
- Qwen3.5-397B-A17B vs Ripple-65B Preview
- Qwen3.5-4B vs Ripple-65B Preview
- Qwen3.5-9B vs Ripple-65B Preview
- Qwen3.6-27B vs Ripple-65B Preview
- Qwen3.6-35B-A3B vs Ripple-65B Preview
- Qwen3.8-27B vs Ripple-65B Preview
- Qwen3.8-Flash-Next vs Ripple-65B Preview
- Qwen3-Coder-30B-A3B-Instruct vs Ripple-65B Preview
- Qwen3-Coder-Next vs Ripple-65B Preview
- Qwen3-Omni-30B-A3B-Instruct vs Ripple-65B Preview
- Qwen3-VL-2B-Instruct vs Ripple-65B Preview
- Qwen3-VL-30B-A3B-Instruct vs Ripple-65B Preview
- Qwen3-VL-32B-Instruct vs Ripple-65B Preview
- Qwen3-VL-4B-Instruct vs Ripple-65B Preview
- Qwen3-VL-8B-Instruct vs Ripple-65B Preview
- Qwopus3.8-27B-Flash-V2 vs Ripple-65B Preview
- Recursive Flint vs Ripple-65B Preview
- Ripple-65B Preview vs SmolLM3-3B
- Ripple-65B Preview vs Spark-X2.6-4B
- Ripple-65B Preview vs Swift 1.5 Qwen3.8-27B
- Ripple-65B Preview vs Swift 1.5 Qwen3.8-Flash-Next
- Ripple-65B Preview vs ThinkingCap (Qwen3.8 27B)
- Ripple-65B Preview vs Baidu Unlimited-OCR-2
Machine-readable
- This page as Markdown
/compare/kronos-vs-ripple-65b-preview.md· slim.min.md· JSON.json(or sendAccept: text/markdown) - Each listing in full
/api/v1/tools/kronos.json·/api/v1/tools/ripple-65b-preview.json - From a terminal
anchor compare kronos ripple-65b-preview(the CLI) - Over MCP
compare_tools {"a": "kronos", "b": "ripple-65b-preview"}at/mcp, no key