
Macroscope Review, Pricing & Alternatives (2026)
Usage based code review, repository intelligence, and coding agents
Last updated: October 26, 2026
What is Macroscope?
Macroscope is a GitHub Cloud application for pull request review, repository questions, engineering summaries, merge checks, and cloud coding agents. It targets software teams that want review automation without per developer seat fees. New workspaces receive $100 in usage credit, Agent includes 1,000 monthly credits, and Code Review starts at $0.25 per review.

Macroscope is a GitHub Cloud application for pull request correctness review, repository questions, engineering summaries, merge checks, and coding agents. The company was founded in 2023 by Kayvon Beykpour, Joseph Bernstein, and Rob Bishop, then publicly launched on September 17, 2025. CNBC reported $40 million in total funding, including a $30 million Series A, while TechCrunch identified Lightspeed, Thrive, Adverb, and GV among the investors. Its audience is software teams that want automated review and repository context without paying for every developer seat. Macroscope supports GitHub Cloud as a GitHub App, rather than GitLab, Bitbucket, or a self hosted repository service. The company also describes customer code isolation, encryption in transit and at rest, and SOC 2 Type II controls for team deployments.
A typical workflow begins with a GitHub pull request. Macroscope evaluates the unified diff for correctness issues, posts review findings, and can offer suggested fixes. Four Detection Modes trade review cost and coverage: Budget, Balanced, Precise, and Ultra. Status summarizes commits and engineering activity, while Ask Macroscope answers questions about code and history. Check Run Agents provide customizable repository checks, and Approvability evaluates whether a pull request meets merge conditions. The CLI reviews local changes, and Murmur runs cloud coding agents in isolated sandboxes. Teams can configure review triggers, ignored files, repository controls, per review caps, per pull request caps, and monthly spending limits. Agent credits cover queries, macros, webhooks, fixes, CLI work, and Check Run Agents.
Macroscope combines model backed services rather than presenting one named foundation model as its product identity. Its pricing documentation says Agent charges consist of raw model usage plus a 5 percent markup, and its security page names OpenAI and Anthropic as subprocessors. The company says it does not train models on customer source code, while its model provider agreements prohibit provider training on that code. Product releases from 2025 and 2026 added Approvability, usage based pricing, Agent pricing, Check Run Agents, open source model support for those agents, Detection Modes, and Murmur cloud coding agents. The July 2026 release added CLI support and open source model support for Check Run Agents. That option applies to selected agents, not to Macroscope itself, which remains proprietary.
Macroscope uses a prepaid shared workspace balance and does not charge seat fees. New workspaces receive $100 in nonexpiring usage credit, while each workspace receives 1,000 Agent credits monthly, worth $10. Those monthly credits do not roll over. Code Review costs $0.025 per KB in Budget mode, $0.05 in Balanced mode, $0.06 in Precise mode, or $0.20 in Ultra mode. Minimum review charges are $0.25, $0.50, $0.60, and $2, with a 10 KB minimum diff. Status costs $0.05 per successfully summarized commit. Approvability costs 30 credits when its eligibility agent runs, while deterministic decisions are free. Agent work costs raw model usage plus 5 percent. Automatic refill is enabled by default, and workspace spending limits can constrain ongoing usage.
Macroscope's main limitation is its GitHub Cloud requirement, which excludes self hosted deployments and GitLab or Bitbucket repositories. Usage billing can also make totals harder to forecast when reviews, summaries, agents, and coding work share one balance, although per review, pull request, and monthly limits reduce that risk. CodeRabbit focuses more narrowly on pull request review, Cursor centers on an AI coding editor, and GitHub Copilot covers coding assistance across Microsoft's developer tools. Qodo adds review and software quality workflows, while Greptile emphasizes repository understanding and review. Macroscope stands apart through Status reporting, Approvability, Check Run Agents, and usage based review pricing. GitHub centered teams needing review automation plus repository operations should pick Macroscope.
Who is Macroscope best for?
- It suits gitHub teams automating pull request correctness checks.
- It suits engineering leaders tracking commits and development activity.
- It suits teams needing merge eligibility decisions alongside code review.
- It suits organizations using cloud coding agents with spending controls.
Macroscope Pricing (2026)
Pricing model: Usage-based. All prices in USD.
| Plan | Price | What you get |
|---|---|---|
| New workspace credit | $0 | Includes $100 in usage credit. The balance does not expire and applies across Code Review, Status, and Agent usage. |
| Agent monthly allowance | $0 | Includes 1,000 Agent credits per workspace each month, worth $10. Credits do not roll over. |
| Code Review, Budget | $0.025/KB | Each review has a $0.25 minimum and a 10 KB minimum diff. This mode has the lowest review cost. |
| Code Review, Balanced | $0.05/KB | The default mode has a $0.50 minimum and a 10 KB minimum diff. Macroscope describes it as broad bug detection. |
| Code Review, Precise | $0.06/KB | Each review has a $0.60 minimum and a 10 KB minimum diff. Agentic validation is added to reduce false positives. |
| Code Review, Ultra | $0.20/KB | Each review has a $2 minimum and a 10 KB minimum diff. It uses more parallel agents and advanced models. |
| Status | $0.05/commit | Charges apply to each successfully summarized commit. Each commit is summarized at most once. |
| Agent usage | $0.01/credit | Credits cover raw model usage plus a 5 percent markup for Check Run Agents, CLI reviews, queries, macros, webhooks, and fixes. |
| Approvability | $0.30/run | An eligibility agent run costs 30 credits. Deterministic eligibility decisions are free. |
| Enterprise | Custom | Enterprise plans can include custom credit packages, annual commitments, priority support, an MSA, and a DPA. |
Pros and Cons
Pros
- It usage pricing avoids per developer seat fees.
- It four review modes let teams trade cost for analysis.
- It repository summaries and merge checks extend beyond comments.
- It sOC 2 Type II and code isolation support team deployments.
Cons
- It gitHub Cloud is the only supported hosting platform.
- It shared usage meters make total costs harder to predict.
- Macroscope itself is proprietary and not self hostable.
Who should look elsewhere?
- It is a poor fit for teams hosting repositories on GitLab or Bitbucket.
- It is a poor fit for organizations requiring a self hosted review service.
- It is a poor fit for small projects seeking a fixed monthly subscription.
Technical Details
Frequently Asked Questions
How much does Macroscope cost?
Macroscope uses usage based pricing with no seat fees. Code Review ranges from $0.025 to $0.20 per KB, depending on Detection Mode, with minimum charges from $0.25 to $2. Status costs $0.05 per summarized commit. Agent usage costs $0.01 per credit, covering raw model usage plus a 5 percent markup. New workspaces receive $100 in usage credit.
Does Macroscope have a free plan?
Macroscope does not present a conventional unlimited free plan. Each new workspace receives $100 in usage credit, and each workspace receives 1,000 free Agent credits per month. The welcome balance does not expire, while monthly Agent credits do not roll over. Teams can also set per review, per pull request, and monthly spending limits.
How does Macroscope compare with CodeRabbit?
CodeRabbit is primarily an AI pull request review product, while Macroscope adds Status summaries, Ask Macroscope, Approvability, Check Run Agents, CLI reviews, and Murmur cloud coding agents. Macroscope also bills by usage rather than seats. Teams wanting broader GitHub repository operations may prefer Macroscope, while teams seeking a narrower review workflow may prefer CodeRabbit.
Does Macroscope train on customer code?
Macroscope says it does not train models on customer source code, and says its model provider agreements prohibit provider training on that code. The company lists encryption in transit and at rest, customer code isolation, SOC 2 Type II, and subprocessors including GCP, OpenAI, Anthropic, and Slack. Macroscope supports GitHub Cloud rather than self hosted deployment.
Can teams use Macroscope commercially?
Macroscope is designed for commercial software teams and offers enterprise agreements with custom pricing, an MSA, a DPA, custom credit packages, annual commitments, and priority support. Its terms govern customer content and service use. The company says customer content is deleted within 30 days after cancellation or termination, subject to backups, legal holds, and legal retention requirements.
Best Macroscope Alternatives in 2026
These tools do the same job as Macroscope, with different pricing, limits and trade-offs.
AI coding workspace with repository context and software agents
AI coding assistance across editors, terminals, and GitHub

Context-aware AI review for pull requests and coding workflows

AI code review and governance for teams shipping agent-generated software

Repository-aware AI review for pull requests and coding agents