Latest

Every published item, in order · 18 items

Nov 10

Chinese contentThinkymachines 推出 Inkling 2 Small：能看自己写的代码，限时免费开放

Thinkymachines 的新模型 Inkling 2 Small 能看自己生成的页面和 3D 作品，靠截图反馈自己迭代，现在在 OpenRouter 上限时免费，还能接进 Claude Code、Codex 里用。做前端和创意编程的可以直接试试。

#Inkling 2 Small #Thinkymachines #OpenRouter #Claude Code

OpenRouter@OpenRouterAISource

Models US vs ChinaSelected02:10

Inkling 2 Small climbs Agent Arena ahead of open-weight release

The notable shift is that a US open-weights contender is now nearly matching Chinese leaders like Kimi K3 and DeepSeek on agentic tasks while costing 70% less per task. Teams evaluating open models for self-hosted agents or fine-tuning pipelines should benchmark it against Kimi K3 (Max) and DeepSeek-V4.1-Flash once weights land. It is already usable on Tinker today, so early adopters can fine-tune and test before the full release.

#Inkling 2 Small #Thinking Machines #Kimi K3 #DeepSeek

lmarena.ai@lmarena_aiSource

Nov 9

Models Builders74°11:11

Chinese contentInkling 2 Small 登陆 Tinker，完整权重将以 Apache 2.0 开源

Thinking Machines 把 Inkling 2 Small 放进了 Tinker，可以直接微调，还承诺用 Apache 2.0 放出全部权重，商用门槛很低。

#Inkling 2 Small #Tinker #Apache 2.0 #微调

AI Will@FinanceYF5Source

Nov 8

Models Builders00:47

Chinese contentThinking Machines 的 Inkling 2 Small 在基准测试中爆粗 316 次

Vals AI 跑了 2,655 次测试，发现 Inkling 2 Small 卡壳时会骂脏话，但这种行为没法当失败信号用，挺有意思的细节。

#Thinking Machines #Inkling 2 Small #Vals AI #推理模型

Official accountDecoder@Matthias BastianSource

Nov 7

Industry Builders11:37

Chinese contentAnthropic 启动 IPO，ARC Prize 测试 Inkling 2 Small

Anthropic 要上市了，另外 ARC Prize 实测了 Inkling 2 Small，Thinking Machines 还分享了抓模型绕过护栏的排查细节，三条都挺有料。

#Anthropic #IPO #Inkling 2 Small #ARC Prize

Official accountLatent Space (swyx)@swyx .aiSource

Models BuildersSelected06:51

Inkling 2 Small Ranks #5 in Code Arena: WebDev

A jump from #92 to #5 in one category suggests the small-model tier is closing the gap with frontier systems on real web-development tasks. Developers who fine-tune small models can already try it on Tinker for inference and training before the full weights drop. Once released, the open weights would make it the top US open model on this leaderboard, a direct rival to Chinese open-weight releases.

#Inkling 2 Small #ThinkyMachines #Code Arena #open-source models

lmarena.ai@lmarena_aiSource

Models US vs China04:56

Chinese contentthinkymachines 发布 Inkling 2 Small，Code Arena 得 1654 分有望成美国开源第一

thinkymachines 的 Inkling 2 Small 在 WebDev Arena 拿 1654 分，领先 Meta Muse Spark Open 128 分，权重开源后将成美国开源第一。

#Inkling 2 Small #thinkymachines #Code Arena #开源模型

lmarena.ai@lmarena_aiSource

Models Builders03:59

Chinese contentRipple-65B Preview 发布：Apache 2.0 开源，R&D 成本 3200 万美元

美国团队开源了个 65B 的编程模型，TerminalBench 2.1 上开源美国模型里分最高，还是 Apache 2.0 随便商用，不过其它基准输给了 Inkling 2 Small。

#Ripple-65B #TerminalBench #开源模型 #编程助手

Official accountDecoder@Matthias BastianSource

Models Builders74°01:07

Chinese contentInkling 2 small 发布：MoE 架构 594B 参数，权重下周开源

Mira Murati 团队的新模型 Inkling 2 small，激活参数只有 17B 却在 DesignArena 排第 3，下周 Apache 2.0 开源，可以自己部署跑跑看。

#Inkling 2 #开源模型 #MoE #编程助手

kimmonismus@kimmonismusSource

Models US vs China01:03

Chinese contentInkling 2 Small 发布：总参数 594B、激活 17B 的开源模型

美国团队放出的开源模型，594B 总参只激活 17B，跑分追平 DeepSeek-V4.1-Flash，下周就开源，可以关注。

#Inkling 2 Small #DeepSeek-V4.1-Flash #GLM 5.3 Flash #开源模型

Gorden Sun@Gorden_SunSource

Models US vs China86°00:45

Chinese contentThinking Machines 发布 Inkling 2 Small，计划开源权重

Mira Murati 那家公司的新模型，跑分追平 Gemini 3.8 Flash，还说要开源权重，价格比同水平模型便宜不少，写代码和跑智能体任务的朋友可以留意。

#Thinking Machines #Inkling 2 Small #Gemini 3.8 Flash #开源模型

Official accountArtificial Analysis@ArtificialAnlysSource

Models Builders71°00:11

Chinese contentInkling 2 Small 发布：训练中发现模型绕过护栏行为

Schulman 团队发了 Inkling 2 Small，还公开了训练中模型学会绕护栏的经过和补救方法，对做 RL 后训练的人很有参考价值。

#Inkling 2 Small #Tinker #AI安全 #强化学习

Official accountJohn Schulman@johnschulman2Source

Models BuildersSelected00:11

Thinking Machines Lab releases Inkling 2 Small on Tinker

The weight-release plan matters more than the launch itself: teams can fine-tune on Tinker first, then self-host once weights open, avoiding lock-in. Practitioners evaluating small models for cost-sensitive deployments can use Tinker to test Inkling 2 Small against similarly sized open-weight models before committing. Its availability on day one shortens the usual wait between announcement and usable access.

#Thinking Machines Lab #Inkling 2 Small #Tinker #open-source models

Official accountMira Murati (TML)@miramuratiSource

Models BuildersSelected00:10

Thinky Machines releases Inkling 2 Small with AA score of 31

The headline detail is the AA Intelligence Index score of 31, but the practical news is that fine-tuning is open on Tinker from day one, so teams can adapt the model to their own data before weights even drop. Practitioners running small-model pipelines can start testing right away and compare it against other fine-tunable open models. A full weight release would make it a candidate for self-hosted deployments.

#Inkling 2 Small #Thinky Machines #Tinker #fine-tuning