Meta's terminal coding agent for developers working through long, multi-step coding tasks
Fieldwork Code
Runs several AI coding sessions at once with a choice of agent harness and model, on a laptop or in cloud sandboxes, and can be steered from a phone; local use is free, cloud is $50 per user per month.
- United States
- New York
- Software Development
- Artificial Intelligence - AI
TrustList summary
Not yet independently verified. Dated from the Product Hunt feed (18 November 2026, Pacific time); the maker's own site shows no launch date or announcement. Features and prices are the maker's own statements, read 19 November 2026, and no outside source was read. The address, in New York, is from the maker's site. This product is unrelated to the market-research firm Fieldwork already listed here. We will update this when it can be confirmed, and remove this note.
Developers who use more than one coding agent usually end up with several terminals and no common place to see the work. Fieldwork Code, from Fieldwork AI Inc. of New York, is a coding tool that runs parallel sessions with the harness and model of the user's choice and shows their progress in one view.
How it works
A user connects repositories and starts sessions that run either on their own computer or in a cloud sandbox that keeps running when the laptop is closed. Harnesses include Fieldwork's own agent, Claude Code, Codex, OpenCode and Pi, with models from Anthropic, OpenAI, Google, xAI, Meta, Moonshot AI, Alibaba, Z.ai and DeepSeek. Scripts per conversation set up worktrees, run development servers and clean up on archive. A side pane shows the terminal, file changes, pull-request status and continuous-integration checks, and work can be grouped into team-named stages. Sessions can be started or continued from the web or from mobile apps, and the product sits inside a wider Fieldwork offer with connectors, task tracking and a versioned knowledge base. Desktop apps are listed for Mac and Windows.
Prices
Local use, running on the user's own Claude or ChatGPT plan, is free. The cloud plan is $50 per user per month and includes 1,440 vCPU-hours and 2,880 gigabyte-hours a month, enough for a small sandbox running around the clock. Larger sandboxes are billed at standard rates beyond that.
Key facts
Maker: Fieldwork AI Inc., New York Launched: 18 November 2026 (Product Hunt feed) Product: parallel AI coding sessions, local or cloud Price: local free; cloud $50 per user per month Investor named on site: firstminute capital Website: fieldwork.ai
Categories & features
- United States
- New York
- Software Development
- Artificial Intelligence - AI
Request a demo or quote from Fieldwork Code
Reviews
Write the first review of Fieldwork Code
Used it? Your experience helps other buyers decide.
Write a reviewQuestions & answers
No questions yet. Be the first to ask about Fieldwork Code.
Fieldwork Code alternatives
Similar listings buyers compare against this one.
An open-source workspace where engineering teams coordinate several AI coding agents together
An AI agent that triages and fixes software bugs so engineering teams spend less developer time
Lets engineering teams join one running Codex or Claude Code session to code together
A terminal built around Claude Code for developers, with one-click resume of interrupted sessions
Makes past AI agent and IDE conversations searchable for developers who need to find them
More on TrustList
Everything here links back to the same verified catalogue. Pick your next stop.
- Write a reviewHelp other buyers decide
- CompaniesAgencies, consultancies and IT service providers, ranked by verified reviews.
- ArticlesGuides, comparisons and research from the editorial desk and the community.
- AwardsAnnual recognition decided by verified reviews and an independent jury.
- LaunchesNew products and releases, voted up by the community every day.
- AI ModelsBenchmark scores and community ratings for every major model.
- RequestsBuyers describe what they need; vendors respond directly.
- PeopleReviewers, authors and makers with public profiles.
- ComparePut up to four listings side by side before you shortlist.