Ollama: Remote Cancel Fixes and Wider Model Support
Ollama focused on correctness for proxied and translated requests, with fixes for remote cancellation, reasoning order, and tool calls. Parallel work widens backend support with an MLX drafting proposal, a llama.cpp bump, and improved model detection.
Duration: PT2S
Episode overview
This episode is a short developer briefing from Ollama.
It explains recent repository work in plain language.
- Show: Ollama
- Published: 2026-10-20T13:01:09Z
- Audio duration: PT2S
Transcript excerpt
This excerpt keeps the crawler page concise. Listen to the episode or use the RSS feed for the full update.
Hello, it's Tuesday, October twentieth, twenty twenty-six, and this is your Ollama developer briefing.
The headline pattern is correctness fixes where Ollama translates or forwards requests, with a second push to widen model and backend support.
First, remote and Open AI compatibility. Merged change number nineteen one forty-four now passes the client request context to upstream chat and generate calls for remote models, so disconnecting locally actually cancels the upstream work instead of leaving it running. Related fixes tighten tool calling: number…
Second, inference backends. Proposal nineteen one forty-nine adds D Flash Two speculative drafting for M L X, drafting a fixed seven-token chain for target verification, alongside the proposed llama dot C P P bump to build eleven seven thirty-eight in number nineteen one forty-three. Two smaller detection fixes fit…
Third, usability and maintenance. Open proposal nineteen one fifty-three adds list with a cloud flag to discover runnable cloud models without a daemon or sign-in, while nineteen one fifty-seven would let self-hosted remote models appear correctly in Claude Desktop, and nineteen one forty-eight would warn at launch…
Expect…
Nearby episodes from Ollama
- MLX Compatibility Push and Responses Fixes
- Weekly Recap - Decision Models Go Live, MLX Gains, and API Polish
- Rendering Transparency and Response Fidelity Fixes
- Proxy Transfers and API Resilience
- Model Compatibility Catch-Up and Windows Reliability
- Responses Compatibility and Clearer Errors
- Decision Models Go Usable, MLX Speeds Up on CUDA
- Decision Scoring Matures as Inference Hardens