Ollama 發布 v0.40.17,加入 Inkling 2 Mini 並改善多項功能
v0.40.17
Ollama v0.40.17 加入 Thinking Machines 的 Inkling 2 Mini:它有 26B 個引數、每次運作會用到其中 5.4B,能處理文字和圖片、呼叫工具,也能進行推理;在 SWE-Bench Verified 的分數是 79.2%。這個模型約佔 20GB,能在 Apple silicon 的 Mac 上執行;這次版本還不支援 PC,之後才會加入。這次更新也讓 Apple silicon 上的 Ollama 能同時處理多個請求,並改善 Claude Code 延後載入工具等功能。其他修正包括讓共用記憶體的顯示卡載入模型更順利,以及避免 Ollama 在非互動式終端機中因沒有輸入指令而顯示錯誤。
AI 伴隨問答
針對這篇報導有疑問?可直接點選快捷問題或向 AI 追問背景脈絡與技術細節:
Inkling 2 Mini
Inkling 2 Mini from Thinking Machines is a 26B parameter model with 5.4B active.
Its looped layers give it more computation per token than its size suggests, and it still fits on a Mac at about 20GB. It handles text and images, tool calls, and thinking, and scores 79.2% on SWE-Bench Verified).
ollama run inkling-2-mini
Note: inkling-2-mini is supported on Apple silicon only in this release. PC support will follow in an upcoming release.
What's Changed
-
ollama launch claudesupports Claude Code's deferred tool loading, so MCP tools are loaded when needed instead of all at
once -
Ollama on Apple silicon now serves multiple requests at once.
-
ollama launch opencodesets thinking options from the model's metadata -
ollama launch pirefreshes Pi's model list after sign-in and passes along prompt arguments -
ollama launch chatgpt --restoreasks before switching an active profile and preserves existing settings and chat history;
quitting ChatGPT yourself now lets it exit through its own quit prompt -
Running
ollamawith no arguments in a non-interactive shell prints help instead of erroring -
Ctrl+C or Ctrl+G while in a multiline
"""block leaves multiline mode instead of swallowing the next line -
Changing the context shift setting no longer reloads the model
-
Vulkan now uses mmap for integrated GPUs, fixing
Failed to allocate pinned memoryload failures on shared-memory GPUs -
On Windows iGPUs, partially offloaded models no longer fail to allocate the KV cache
-
Tool definitions passed through the API keep JSON schema keywords Ollama doesn't model instead of dropping them
-
Disabling thinking now maps to the model's lowest supported reasoning level rather than being ignored
-
Source builds can use prebuilt MLX-C and XGrammar binaries
Full Changelog: v0.40.16...v0.40.17-rc3