MiroThinker
MiroThinker is a deep research agent optimized for complex research and prediction tasks. Our latest models, MiroThinker-1.7, achieves 74.0 and 75.3 on the BrowseComp and BrowseComp Zh, respectively.
Runs with mockssource: GitHubhomepagePythonApache-2.0commit 1c4253f6774b
Python, Apache-2.0 licensed. The project labels itself: agent, agent framework, browsecomp, deep research, futurex, gaia, hle and research agent.
MiroThinker runs, with stand-ins for the services it depends on. An Argusic agent installed it in 12 minutes and hit 1 error and fixed every one without help on a clean machine with no GPU, and the whole session was recorded.
All 5 packages install and all source modules import correctly. The agent pipeline executes end-to-end with a mock LLM server: connects to MCP tool servers, calls LLM, processes response, writes task log with status=success. Flask trace visualizer answers HTTP 200.
What the agent ended up with on a clean machine, in its own words. How this is measured
Measured by Argusic on a fresh machine every time. Every number links to its evidence. Argusic Score 92 of 100 (the mean of this project's run scores).
At a glance
Subject data from GitHub, linked at the top of this page, refreshed . Test data by Argusic (CC BY 4.0); every number links to a run page with the full log, the recording, and their sha256 hashes.
What was measured
From the run that decided this verdict. Every line is a recorded value, not an estimate.
What went wrong, in the agent's own words
The agent's own words, worst first, never rewritten into advice.
- miroflow-tools resolved mcp>=2 / fastmcp>=4 (v2 API), but code uses FastMCP from v14 minutes
- Install time
- 12 minutes
- Cold machine to finish
- 12 minutes
- Errors hit and fixed
- 1 hit, 1 fixed with no human help
- How the result was proved
- uv run python main.py agent=demo llm=gpt-5 llm.base_url=http://localhost:19999/v1 connected to mock OpenAI server, completed pipeline in 1 turn with output 'Attention Is All You Need (mock)', logged status=success
- Model tokens used
- 95,524
- Exact commit tested
- 1c4253f6774b
- Container image
- sha256:cdd920bce783
What was not measured
The other half, so you can see exactly how far this evidence goes. A gap is printed as a gap and never filled in.
- Peak GPU memory
- this ran on a CPU machine, so there was no GPU to sample
- Behaviour over a long run
- each attempt uses the smallest documented settings and a fixed time budget, so this measures whether it runs, not how it holds up under load
- Behaviour against real services
- this run was verified against stand-in services only
The end of the recording
The last lines of the terminal session that decided this verdict, exactly as they were recorded.
2026-09-28T01:54:46.813289Z ERROR codex_core::session: failed to record rollout items: thread 01a0e… 2026-09-28T01:54:46.813482Z ERROR codex_core::session: failed to record rollout items: thread 01a0e… tokens used 95,524 Now I have all the information. Let me compute the overall results. The install took place across multiple `uv sync` operations. Let me calculate total time: - `miroflow-agent` first sync: ~6.6s - `miroflow-tools` initial sync: ~3s, then re-sync after fix: ~1s - `gradio-demo` sync: ~6.9s - `visualize-trace` sync: ~1.3s - `collect-trace` sync: ~1.8s Total install time roughly 6-8 minutes (including exploration, learning, and fixing the mcp version…
Replay the whole session, every command from a clean machine to this point.
How it was tested
One agent, one container, no human help. It installs the project the way its own documentation says to, runs it, and fixes what breaks. Everything below is recorded as it happened: the terminal session, the log and the exact commit. The full method.
Strengths and limits
Measured facts, not opinions. How this is written.
What went well
- Recovered from all 1 error without a human stepping in, which says the failures are documented well enough to solve.
- Licensed Apache-2.0, as reported by its host.
What did not, or is not known
- Only reached a running state against stand-in services, so nothing here proves it works against the real ones.
- Took 12 minutes to install, slower than the median of the 48 comparable projects Argusic has measured.
- Measured once so far. A single attempt shows it can run, not that it runs reliably.
Of the 49 agent projects Argusic has installed and timed, MiroThinker was the 30th fastest to reach a running state, and 34 of 49 reached one at all.
Also tested, in the same area
Every one of these was installed and run by Argusic on a clean machine. Nothing appears here that was not tested.
Run history
| tested by | attempt | outcome | score | cost USD | when |
|---|---|---|---|---|---|
| Argusic Runner | 1/3 | Runs with mocks | 92.00 | 0.10 |
Topics (from GitHub)
agentagent-frameworkbrowsecompdeep-researchfuturexgaiahleresearch-agentsearch-agentxbench
Embed the badge
Markdown for the project README. It links back here; terms on the terms page.
[](https://argusic.com/subject/mirothinker)Questions
- Does MiroThinker run?
- MiroThinker runs, with mocks standing in for external services. Argusic installed and launched it on a clean machine in 12 minutes, hitting 1 error on the way, and recorded the session.
- How did Argusic test MiroThinker?
- On a fresh, disposable machine, with every command recorded and the repository pinned at commit 1c4253f6774b. One attempt is recorded, and the full method is on the methodology page.
- What did it cost to test MiroThinker?
- The run that produced this verdict cost $0.10: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
- How long does MiroThinker take to install?
- 12 minutes in the run that decided this verdict, from a clean machine to a running state. That is one measured attempt, not an average.
- Does MiroThinker need a GPU?
- Not to start. Argusic reached a running state on a machine with no GPU.
- What goes wrong when installing MiroThinker?
- 1 thing broke in the recorded run, and 1 were fixed without human help. Each one, and the time it cost, is listed on this page.
- How does MiroThinker compare with the alternatives?
- Of the 49 agent projects Argusic has installed and timed, MiroThinker was the 30th fastest to reach a running state, and 34 of 49 reached one at all.
- Where is the evidence for MiroThinker?
- The recorded run is on this page, each linking to its full log and terminal recording, stored with a sha256 fingerprint so it cannot be quietly altered.