Xiaomi's new open source, agentic AI coding harness Mi Mo Code beats Claude Code at ultra-long, 200+ step tasks | Venture Beat

Overview

Xiaomi's new open source, agentic AI coding harness Mi Mo Code beats Claude Code at ultra-long, 200+ step tasks

Xiaomi's Mi Mo AI team has open-sourced Mi Mo Code V0.1.0, a terminal-native AI coding assistant that the Chinese electronics giant says outperforms Anthropic's Claude Code on key agentic coding benchmarks, especially on long-horizon, multi-step tasks (200+ steps) — at least, according to its own internal beta release and survey of 576 developers.

Details

The release was announced June 10, 2026 in a post on the social network X from the official @Xiaomi Mi Mo account, which described the tool as "more than an AI coding assistant in your terminal — it's the smartest coding partner you'll ever work with."

Mi Mo Code is available now on Git Hub under an MIT license, and installs with a single terminal command (curl -fs SL https://mimo.xiaomi.com/install | bash) on mac OS and Linux or via npm (npm install -g @mimo-ai/cli) on Windows.

The project is a fork of the open-source Open Code agent, which Xiaomi has extended with its own memory architecture, workflow modes, and model harness.

As any avid vibe coder would surely attest, AI coding agents degrade over long working sessions: as the context window fills, earlier decisions, conventions, and task state get compacted away or lost entirely, forcing developers to re-explain their projects.

Xiaomi argues this approach is doomed at scale. "What we need is not better compression, but an explicit storage-and-retrieval mechanism that decides what information should be written into persistent structures, and when it should be recalled," the Mi Mo team noted in their launch blog.

Mi Mo Code attacks this with a cross-session memory system, powered under the hood by SQLite FTS5 full-text search, that spans four layers: project memory (a persistent MEMORY.md file), session checkpoints, scratch notes, and per-task progress logs.

The note-taking is key, here: Rather than forcing the primary coding agent to pause its work to take notes, the system deploys an independent "checkpoint-writer" subagent.

Think of it the primary coding agent as a construction contractor working to build a massive mansion alongside a dedicated architect, the checkpoint-writer subagent. While the main agent focuses on building out the physical structure, the subagent updates the blueprints in real time, noting decisions, issues, and the actual lay of the land as the construction project progresses.

When the context window approaches its limits — the contractor gets lost in the half-built mansion — it can consult the subagent and find its place again. In the case of Mi Mo Code, the system simply rebuilds the environment from structured checkpoints with the relevant context, ensuring no loss of operational momentum.

Two self-improvement mechanisms round out the system: a /dream command that periodically (roughly every seven days) reviews historical sessions, deduplicates them, and compresses them into long-term memory, and a "distill" function that mines past sessions for repeated workflows that can be automated, following a similar approach taken recently by Open AI and Anthropic with their various models.

Impressive performance on software engineering (SWE) benchmarks

According to benchmark figures published in Xiaomi's technical blog post, Mi Mo Code paired with Mi Mo-V2.5-Pro outperformed Claude Code paired with Claude Sonnet 4.6 on all three evaluations tested:

Mi Mo Code vs. Claude Code benchmark performance. Credit: Xiaomi

Mi Mo Code vs. Claude Code benchmark performance. Credit: Xiaomi

The harness itself accounts for a measurable share of the gain. Running the same Mi Mo-V2.5-Pro model in both harnesses, Mi Mo Code scored 62% on SWE-bench Pro versus 57% for Claude Code, and 73% on Terminal Bench 2 versus 68% — roughly five points each, attributable purely to the agent system rather than the model.

Xiaomi notably did not publish comparisons against Open AI's Codex or Google's Gemini CLI — Claude Code is the sole named competitor throughout its materials, a telling choice of benchmark target.

Independent reference points suggest why. On the official Terminal-Bench 2.0 leaderboard maintained at tbench.ai, Open AI's Codex CLI running GPT-5.5 scores 82.2% — roughly nine points above Mi Mo Code's self-reported 73% — and Open AI's own GPT-5.5 announcement claims 82.7% on the same benchmark.

On SWE-Bench Pro, however, the picture flips: Open AI reports GPT-5.5 at 58.6%, below Mi Mo Code + Mi Mo-V2.5-Pro's claimed 62%. (Mi Mo Code does not yet appear on either official leaderboard, and cross-comparing self-run numbers against leaderboard submissions carries the usual configuration caveats.)

Perhaps more interesting than the offline benchmarks: Xiaomi says it ran a human double-blind A/B evaluation during its internal beta, covering 576 developers working in 474 real private repositories, producing 1,213 judged head-to-head pairs against Claude Code using the same target model.

Under 200 execution steps, the two systems split roughly 50/50 — but past 200 steps, Mi Mo Code's win rate rose above 65%, supporting the company's thesis that its memory and state-management architecture pays off specifically on long-horizon work.

Xiaomi itself concedes the standard benchmarks "still measure one-shot problem-solving ability" and don't capture the tool's multi-session design goals.

As always, these are vendor self-reported numbers that haven't been independently verified, and head-to-head harness comparisons are sensitive to configuration. But the claims are consistent with a broader industry pattern: scaffolding and harness engineering are becoming as important as raw model capability in agentic coding performance.

Easy integration with existing developer systems and voice control

From a user experience standpoint, Mi Mo Code is designed to live where developers already work. It operates directly in the terminal, reading and writing files, running commands, and managing Git.

Out of the box, the tool requires zero configuration, connecting automatically to "Mi Mo Auto"—a free-for-a-limited-time channel powered by Xiaomi’s multimodal Mi Mo V2.5 model, which boasts a massive million-token context window. For developers migrating from existing environments, the transition is frictionless: Mi Mo Code automatically imports MCP servers, custom skills, and API configurations from Claude Code.

Compose mode: Pressing Tab switches the agent into a specification-driven workflow in which the developer describes a high-level goal and the system autonomously executes the full development cycle — design, planning, coding, testing, and review — following what Xiaomi describes as a "heavy planning upfront, stable verification later" strategy.

Compose mode: Pressing Tab switches the agent into a specification-driven workflow in which the developer describes a high-level goal and the system autonomously executes the full development cycle — design, planning, coding, testing, and review — following what Xiaomi describes as a "heavy planning upfront, stable verification later" strategy.

Voice control: Built on Xiaomi's Mi Mo-ASR speech recognition with Ten VAD voice activity detection, developers can dictate and modify instructions verbally and speak commands like "send" and "execute" for fully hands-free operation (available for logged-in users).

Voice control: Built on Xiaomi's Mi Mo-ASR speech recognition with Ten VAD voice activity detection, developers can dictate and modify instructions verbally and speak commands like "send" and "execute" for fully hands-free operation (available for logged-in users).

According to Xiaomi, the gains from the agent harness itself are measurable. Running the same underlying Mi Mo model in both harnesses, the company says Mi Mo Code scored 62% on SWE-Bench Pro versus 57% for Claude Code, and 73% on Terminal Bench 2 versus Claude Code's 68% — roughly five percentage points better on each, attributable purely to the agent system rather than the model.

As always, these are vendor self-reported numbers that haven't been independently verified, and head-to-head harness comparisons are sensitive to configuration. But the claim is consistent with a broader industry pattern: scaffolding and harness engineering are becoming as important as raw model capability in agentic coding performance.

The bigger lure for many developers may be what's bundled in.

Mi Mo Code ships with "Mi Mo Auto," a zero-configuration channel offering free, limited-time access to Mi Mo-V2.5 — the natively multimodal model Xiaomi released in late April 2026, a sparse mixture-of-experts design with 310 billion total parameters (just 15 billion active per inference) and a 1 million token context window, which the company positions as matching Anthropic's Claude Sonnet 4.6 in multimodal agentic work.

As Venture Beat reported when the Mi Mo-V2.5 family launched in April, the models are MIT-licensed and among the most efficient and affordable available for agentic tasks.

The larger Mi Mo-V2.5-Pro — a 1.02-trillion-parameter mixture-of-experts model with 42 billion active parameters and a hybrid-attention architecture — led the open-source field on Xiaomi's Claw Eval agentic benchmark with a 63.8% success rate while consuming only about 70,000 tokens per trajectory, roughly 40–60% fewer than Anthropic's Claude Opus 4.6, Google's Gemini 3.1 Pro, or Open AI's GPT-5.4 needed for comparable results.

Notably, the V2.5-Pro's post-training was explicitly designed to instill "harness awareness" — training the model to manage its own memory and context within agent scaffolds like Claude Code or Open Code — making a Xiaomi-built harness optimized around that capability a logical next step.

Pricing is similarly aggressive: Mi Mo-V2.5 starts at

Venture Beat Frontier AI Model API Pricing Snapshot

For developers who don't want Xiaomi's models at all, Mi Mo Code also supports third-party backends — including token plans from Deep Seek, Moonshot's Kimi, and Zhipu's GLM — along with any Open AI-compatible API, mirroring the bring-your-own-model flexibility of its Open Code parent.

Mi Mo Code lands in an increasingly crowded field of terminal-based coding agents: Anthropic's Claude Code, Open AI's Codex CLI, Google's Gemini CLI, and open-source players like Open Code and Aider.

What's new is the entrant. Xiaomi — the world's third-largest smartphone maker, with a fast-growing EV business — has been methodically building its Mi Mo AI division since the release of the Mi Mo-7B reasoning model in April 2025, following with the Mi Mo-VL vision-language series, Mi Mo-V2-Flash, the 1-trillion-parameter Mi Mo-V2-Pro in March 2026, and the V2.5 flagship family in April.

The effort is led by Fuli Luo, a veteran of Deep Seek's disruptive R1 project, who has characterized Xiaomi's frontier push as a "quiet ambush" — and backed it with a 100-trillion free token grant for builders announced alongside the V2.5 launch.

The playbook is familiar from Deep Seek, Alibaba's Qwen, Mini Max, and Moonshot AI's Kimi series: release genuinely capable models and tooling under permissive licenses at a fraction of U. S. lab pricing, and convert the resulting developer mindshare into a durable ecosystem.

By pairing an open-source agent harness with a free frontier-class model, Xiaomi is effectively eliminating both the licensing and the usage cost of entry — at least for now.

What it means for enterprises and technical decision-makers

For engineering leaders, Mi Mo Code is a low-risk, potentially high-value evaluation candidate: MIT-style licensing permits modification and commercial integration, the Open Code lineage means the architecture is inspectable, and the bring-your-own-model support means it can be pointed at an internally approved endpoint rather than Xiaomi's cloud.

The persistent memory system addresses a real and widely felt pain point in agentic development workflows — one that competitors are also racing to solve.

The countervailing considerations: the "free for a limited time" model access is by definition temporary and routes code context through Xiaomi's servers, which will be a non-starter for organizations with strict data-residency or IP policies; the benchmark edge over Claude Code is self-reported; and a V0.1.0 release number signals exactly what it suggests about maturity.

Teams subject to U. S. government procurement restrictions on Chinese technology vendors should also weigh that context before adopting.

Deep insights for enterprise AI, data, and security leaders

By submitting your email, you agree to our Terms and Privacy Notice.

Key Takeaways

• Xiaomi's new open source, agentic AI coding harness Mi Mo Code beats Claude Code at ultra-long, 200+ step tasks

• Xiaomi's Mi Mo AI team has open-sourced Mi Mo Code V0

• The release was announced June 10, 2026 in a post on the social network X from the official @Xiaomi Mi Mo account, which described the tool as "more than an AI coding assistant in your terminal — it's the smartest coding partner you'll ever work with

• Mi Mo Code is available now on Git Hub under an MIT license, and installs with a single terminal command (curl -fs SL https://mimo

• The project is a fork of the open-source Open Code agent, which Xiaomi has extended with its own memory architecture, workflow modes, and model harness