AITOPAI热报
Models84°

Thinking Machines 的 inkling 2 small 上线 Together AI

Official accountTogether AI@togethercompute
Post· 2 sources
View the post
Why we picked it

Mira Murati 的 Thinking Machines 把 inkling 2 small 放上了 Together AI，594B 参数只激活 17B，还有 1M 上下文。写代码、跑长任务、看图都能覆盖，推理成本却按小模型算。做 agent 或长文档处理的开发者可以直接去 Together AI 试。

Thinking Machines 的 inkling 2 small 已在 Together AI 平台上线。

Body

该模型是一个多模态 MoE 架构，总参数量 594B，每次推理激活 17B，上下文窗口最高支持 1M token。模型面向编程、长程 agentic 任务和视觉三类场景设计。

MoE 架构意味着 594B 的总参数中只有 17B 参与每次前向计算，推理成本接近小模型，但知识容量保持大模型水平。上线 Together AI 后开发者可直接通过该平台的 API 调用，无需自建推理集群。

Original · Together AI

inkling 2 small from thinking machines is now live on together ai

the multimodal moe is built for coding, long-horizon agentic work + vision, with 594b total parameters, 17b active + up to 1m context

try it now: https://t.co/dvSKbIifWU https://t.co/HSeiRfn0pH

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include inkling 2 small, Thinking Machines, Together AI, MoE.

Source attribution: curated from Together AI (@togethercompute). Cross-source corroboration: 2 independent sources have covered this development (including Thinking Machines Lab), confirming industry-wide signal strength.

  • Thinking Machines Lab20:50Source
#inkling 2 small#Thinking Machines#Together AI#MoE#多模态

Keep reading

Recent stories with the same tag or section
ModelsBuildersSelected05:10
Thinky Machines releases Inkling 2 Small weights on HuggingFace

License terms, not model size, are the real differentiator here: Apache 2.0 lets enterprises ship commercial products without the compliance overhead that Llama's custom license imposes. Teams evaluating self-hosted models for cost control or data privacy can fine-tune and deploy Inkling 2 Small with no royalties and no obligation to release source code. Engineering leads comparing open-weight options should shortlist it alongside other permissively licensed releases rather than treating all "open" models as equivalent.

#Inkling 2 Small#Thinky Machines#HuggingFace#open-source models
Official accountThinking Machines Lab@thinkymachinesSource
ModelsBuildersMulti-source04:36
Chinese contentHedge fund 用心理分析模型 Thorsten-4 分析财报电话会动机

这个思路挺有意思：4 年追踪 7,000 多人、标注 3,500 万词训出来的心理分析模型，能从财报电话会里读出高管的动机，已经有基金拿它跑 5,000 万美元的仓位。做量化或者另类数据分析的可以看看。

#Thorsten-4#Large Psychology Model#财报电话会#另类数据
Event
rohanpaul_ai@rohanpaul_aiSource
ModelsBuildersSelected04:30
Retell tops $100M run rate, launches Tell 1 voice model

A voice-AI vendor shipping its own foundation model signals a shift from reselling general-purpose LLMs to owning the speech-to-reasoning stack, which could compress latency and cost for high-volume call operations. Contact-center engineering and CX teams evaluating voice-agent platforms should benchmark Tell 1 against general models like GPT-4o on latency and failure rates. Conductor's automated test-case generation also gives teams a concrete workflow for continuous agent quality control.

#Retell#Tell 1#voice agents#Conductor
TestingCatalog@testingcatalogSource
ModelsUS vs ChinaMulti-source04:20
Chinese contentGoogle Gemini 4 Hydrogen 发布四周即超越自家更大的 Argon

Google 把 Flash 系列换成了 Gemini 4 Hydrogen，结果这个小模型上线四周就在 Artificial Analysis 榜单上超过了自家更贵的 Argon。它和 Claude Sonnet 5.5 得分差不多还更便宜，但比 GPT-6.1 Sol 贵 2.5 倍、写智能体代码不行。做编程或智能体的开发者可以对比一下再选。

#Gemini 4 Hydrogen#Google#Claude Sonnet 5.5#GPT-6.1 Sol
Event
Official accountDecoder@Matthias BastianSource