MiniMax M2 arrives on Ollama’s cloud for coding and agentic workflows

MiniMax M2 is now available on Ollama’s cloud. The source describes it as a model built for coding and agentic workflows. You can start it with `ollama run minimax-m2:cloud`. The page also lists integrations with coding tools including VS Code, Zed, and Droid, and says cloud models such as `minimax-m2:cloud` can be accessed directly via Ollama’s cloud API.

According to benchmarks from Artificial Analysis, MiniMax-M2 demonstrates highly competitive general intelligence across mathematics, science, instruction following, coding, and agentic tool use. Its composite score ranks #1 among open-source models globally, per the source. For coding, the source says it excels at multi-file edits, coding-run-fix loops, and test-validated repairs, with strong performance on Terminal-Bench and (Multi-)SWE-Bench–style tasks. For agent work, it plans and executes complex, long-horizon toolchains across shell, browser, retrieval, and code runners. In BrowseComp-style evaluations, it consistently locates hard-to-surface sources, maintains traceable evidence, and gracefully recovers from flaky steps.

The design is efficient: 10 billion activated parameters (230 billion in total), which the source says delivers lower latency, lower cost, and higher throughput for interactive agents and batched sampling. That matters for people running agents repeatedly or batching requests. The source does not provide specific latency or price figures, and it does not name a previous option for direct comparison. So the practical answer to what this can do that a previous option could not is not fully spelled out in the evidence. What the source does claim is a combination of coding repairs, multi-file edits, and long-horizon tool use in one model available through Ollama’s cloud.