Ollama 0.32: Local Model Runner or AI Agent?
For most of its life, Ollama had one job on this site: it was the quiet backend. You pulled a model, you pointed something at the API, and the interesting decisions happened somewhere else. Version 0.32, released on 11 July 2026, changed the front door.
Type ollama with no arguments now and you do not get help text. You get an interactive agent that offers to chat, write code, search the web, and take on delegated work. The banner it prints names the model it intends to use, and on a default install that model is a hosted one.
That last detail is the whole story, so let us take it slowly.
What actually changed in 0.32.0
Four things, per the release notes:
- The bare
ollamacommand launches an interactive agent. The startup banner advertises chat, code, web search, and delegated work. - The Codex App integration was renamed to ChatGPT. Use
ollama launch chatgpt, with--restoreto return to your usual ChatGPT profile. - The
ollama launchmenu was simplified to the most popular integrations. The others remain reachable throughollama launchdirectly. - Older agent models now warn before launching. CodeLlama, Qwen2.5 and Qwen2.5-coder, the Llama 3.x line, Mistral, StarCoder, and the base DeepSeek-R1 tags all raise a deprecation warning first.
What is conspicuously absent from that list matters as much as what is on it. The release notes record no change to ollama run, and no change to any other model-running subcommand. If your workflow is a script, a systemd unit, a Docker container, or anything else that calls a subcommand explicitly, nothing in 0.32.0 was aimed at you. The new behaviour attaches to the bare invocation only.
The part worth pausing on
The agent banner in the official 0.32.0 notes names glm-5.2:cloud as the model behind the new experience. The :cloud suffix is not decoration. It denotes a model served from Ollama's hosted infrastructure rather than from your GPU.
Version 0.32.1, five days later, added a further tell: agent web search and web fetch now instruct the user to run ollama signin when authentication is required. A local model runner does not need you to sign in. An account-backed cloud service does.
None of this is hidden, and none of it is a scandal. Ollama has been routing some requests to hosted infrastructure for a while, and the company is entitled to build an agent product. But it does change the sentence this site has been able to write for two years. "Install Ollama and your prompts stay on your machine" was a safe default claim. It is no longer a safe default claim about the default experience. It remains true of ollama run against a locally pulled model, which is a narrower statement and needs to be written as the narrower statement.
If you want the old assumption to hold, the practical guidance is unchanged in shape but now needs stating explicitly:
- Call subcommands rather than the bare command.
ollama run <model>against a model you pulled locally is a local inference path. - Pin models locally rather than accepting a cloud-tagged default.
- Treat any agent feature that asks you to sign in as an off-device feature, because that is what the sign-in is for.
The 0.32 point releases, briefly
The series moved fast, and one release should be skipped outright.
- 0.32.1 (16 July). Improved Gemma 4 tool calling and multi-turn reasoning, including more reliable continuations after a tool response. Fixed a recurring MLX model cache leak that could grow memory use across requests. MLX text model loading now respects
OLLAMA_LOAD_TIMEOUT. The agent now receives the current working directory for project context. - 0.32.2 (20 July). Withdrawn. Use 0.32.3 or newer.
- 0.32.3 (23 July). Fixed model downloads that stalled before sending data. Restored Claude Code channels and fixed Anthropic thinking streams. Added CUDA on Windows ARM64 and B200 support through CUDA 12, and lowered memory use on Linux CUDA and ROCm integrated GPUs. Added chat, thinking, and tool calling for Laguna 2.1.
- 0.32.4 (25 July). Laguna on Apple GPUs through MLX. Fixed Qwen3 MoE decoding for differently quantized experts, with a faster packed gate and up projection that Ollama reports at roughly 4 to 9 percent on an M5 Max.
- 0.32.5 (27 July). Fixed an MLX Metal bug that could reduce output quality for NVFP4 models.
- 0.32.6 (4 August). Qwen3.5 is faster on Apple GPUs, because the MLX engine now uses the model's MTP head for speculative decoding automatically. OpenAI-compatible streaming was brought in line with the upstream wire format.
ollama run kimi-k3now offers the cloud tag rather than failing for a cloud-only model that publishes no default tag.
One regression deserves a callout rather than a bullet. 0.32.6 temporarily removed experimental image generation. Ollama's own note says to stay on 0.32.5 if you need it. If image generation is part of your setup, 0.32.6 is not an upgrade.
About that Apple Silicon speed claim
You will see a figure circulating that Ollama generates Gemma 4 tokens nearly 90 percent faster on Apple Silicon, averaged across a coding-agent benchmark, by leaning on multi-token prediction. That number is Ollama's published result.
Two caveats. First, it belongs to version 0.31.1, released 30 June 2026, not to the 0.32 series, and it has been widely misattributed. Second, it is a vendor-reported result from a vendor-selected benchmark. LocalLLMGuide.com has not reproduced it. We report it as Ollama's claim, because that is what it is.
The mechanism is at least easy to describe: Ollama auto-tunes how many tokens to draft as it runs, so the acceleration is on by default, needs no configuration, and Ollama states it does not change model output.
So which is it?
Both, and that is the honest answer rather than a hedge.
Ollama in August 2026 is still the same local inference engine underneath. The MLX and llama.cpp work across the 0.31 and 0.32 series is exactly the sort of unglamorous backend improvement that made it worth recommending in the first place. If you use it as a backend, it is a better backend than it was in May.
What changed is that the project now also ships an agent product, and the agent product is the thing you meet first. For a general developer audience that is a reasonable default. For an audience that came to Ollama specifically because it was the boring local option, the default now points somewhere else, and the burden is on the user to steer away from it.
That is worth knowing before you upgrade. It is not a reason to avoid upgrading.
Related reading
*Version numbers and release contents verified against the official Ollama GitHub releases on 9 August 2026. Performance figures are vendor-reported and have not been independently reproduced by LocalLLMGuide.com.*