Ollama now runs Inkling 2 Mini, and Macs serve several chats at once

1 day ago

Ollama v0.40.16 and v0.40.17 add Inkling 2 Mini (Apple silicon only for now) and Ripple Preview, let Ollama on Apple silicon serve several requests at once, cut Qwen3.8 Flash Next's memory by about 27 GiB, load MCP tools on demand in ollama launch claude, and fix model loading on integrated GPUs.

Ask about this presentation

Chapters

  1. 0:00Ollama now runs Inkling 2 Mini, a 26B model that fits on a Mac
  2. 0:14Ollama on Apple silicon now answers several requests at once
  3. 0:24Ripple Preview, a 65B model for agentic coding, runs with one command
  4. 0:33Qwen3.8 Flash Next now fits on a 128 GB Mac
  5. 0:47ollama launch claude loads MCP tools only when Claude Code needs them
  6. 0:58Integrated GPUs load models that used to fail, and long API inputs stay whole
  7. 1:12Upgrade now if you run Ollama on a Mac
Show transcript
  1. Ollama now runs Inkling 2 Mini from Thinking Machines, a 26 billion parameter model that fits on a Mac in about 20 gigabytes. It reads images, calls tools, and solves about four in five real GitHub issues in a coding test. For now it needs Apple silicon.

    Slide text
    Ollama
    Ollama
    NOV 9–10, 2026
    Ollama now runs Inkling 2 Mini, a 26B model that fits on a Mac
    5.4B active per token
    ~20 GB on a Mac
    79.2% SWE-Bench Verified
    Apple silicon only in this release. PC support follows.
    Ollama v0.40.17 release notes · PR #19799
  2. On Apple silicon, Ollama now serves several requests at once. A second chat or agent used to wait for the first to finish. Left at the default, Ollama picks up to eight parallel slots from the memory you have free.

    Slide text
    Ollama
    Ollama
    NOV 9–10, 2026
    • Before: one request at a time
    • Now: 1, 2, 4 or 8 slots, picked from free memory
    Ollama v0.40.17 release notes · PRs #19197, #19785
  3. Ripple Preview, from Ricursive Intelligence, is a 65 billion parameter dense model built for agentic coding. Ollama run ripple preview pulls it and starts a session.

    Slide text
    Ollama
    Ollama
    NOV 9–10, 2026
    $ ollama run ripple-preview
    Ollama v0.40.16 release notes · PR #19592
  4. Qwen 3.8 Flash Next now uses about 27 gigabytes less memory on Apple silicon. It drops from roughly 98 to 71, which puts it within reach of a Mac with 128 gigabytes, at about the same speed.

    Slide text
    Ollama
    Ollama
    NOV 9–10, 2026
    Before
    97.6 GiB
    Now
    70.8 GiB
    Wired model memory on Apple silicon, same model files
    Ollama v0.40.16 release notes · PR #18078
  5. Ollama launch Claude now supports Claude Code's deferred tool loading, so MCP tools load when they're needed instead of all at once. Claude Code always routes to the model you selected, and a retired cloud model gets an explanation.

    Slide text
    Ollama
    Ollama
    NOV 9–10, 2026
    • MCP tools load on demand, through Claude Code’s deferred tool loading
    • Requests always go to the model you picked
    • A retired cloud model gets an explanation before launch
    Ollama v0.40.16 and v0.40.17 release notes · PRs #19578, #19587
  6. On integrated GPUs, Vulkan now memory-maps the model, which fixes the failed to allocate pinned memory error. On Windows, partly offloaded models now get room for their cache. And the Responses API stops cutting long inputs short unless you ask it to.

    Slide text
    Ollama
    Ollama
    NOV 9–10, 2026
    • Vulkan uses mmap on integrated GPUs, fixing “Failed to allocate pinned memory”
    • Windows iGPUs allocate the KV cache for partially offloaded models
    • /v1/responses keeps long inputs whole unless you set truncation
    Ollama v0.40.16 and v0.40.17 release notes · PRs #19640, #19636, #19501
  7. Should you upgrade? I would say yes on Apple silicon, where Inkling 2 Mini and parallel requests run. On a PC with an integrated GPU, the loading fixes are reason enough, though Inkling waits for a later release. Ollama's release notes list every change.

    Slide text
    Ollama
    Ollama
    NOV 9–10, 2026
    Ollama v0.40.17 release notes

What happened in AI today, in your inbox

One short email each evening. Read it and you’re caught up.

Built with ❤️ in New York