← November 10, 2026 briefing

Ollama v0.40.16-rc0 fixes Claude Code routing errors, cuts Qwen3.8 Flash Next memory use on MLX, adds CUDA 13.5 for ARM64 Windows

Announced November 9, 2026

Verified (fact)

Ollama released a v0.40.16 release candidate. It fixes routing and deprecated-model errors when running Claude Code, reduces memory use for Qwen3.8 Flash Next in the MLX runner, and adds CUDA 13.5 and SM 103 for ARM64 Windows. The OpenAI-compatible API now passes through the truncation option on responses requests, and /v1/models returns an empty array when no models are present.

Why it matters (interpretation)

These are practical improvements for users running Claude Code with local models and for Apple Silicon users.

Sources

More about Ollama