Amp
Multi-model coding agent with CLI and cloud Orbs for long-running software tasks.
Score calculation
(3 + 5 + 3 + 1 + 7) ÷ 5 = 3.8
5 of six dimensions rated. Unrated dimensions are excluded from the average.
Full scoring method ↗Setup details
Amp CLI and cloud service, including hosted Orbs. Selected code context and thread data are processed through Amp and the chosen inference provider.
What fits your needs
Amp CLI and cloud service, including hosted Orbs. Selected code context and thread data are processed through Amp and the chosen inference provider.
Cloud agent tasks
Orbs provide hosted machines per thread for long-running work.
See the evidence
Model modes
Amp offers selectable model modes and provider connections.
See the evidence
Tradeoffs to expect
Code is processed remotely
Selected snippets, prompts and tool results can pass through Amp and its inference providers.
See the evidence
No self-hosted service
Amp says it does not offer self-hosted deployments.
See the evidence
How to switch
Change client route
A local Pi setup can change the client-side route; review each new model provider separately.
Practical questions
Review provider controls
Check provider retention settings before connecting a personal subscription.
See the evidence
Review thread access
Workspace administrators may access private threads under the documented controls.
See the evidence
Amp documents a cloud service and selectable models, but reports no energy or electricity for coding tasks.
See the evidence
The cloud deployment and model modes define routes but no footprint, electricity sourcing or measured resource use. Mode choices don’t show energy savings.
1 + 1 reporting + 0 electricity + 1 controls = 3/10. This grades documented practices and transparency.
Per-task energy, emissions, water and facility-level clean power are not reported.
- Reporting1/3
- Amp documents its cloud service and provider boundary but no product energy report.
- Electricity sourcing0/3
- No clean electricity or facility mix is reported.
- Resource controls1/3
- Selectable model modes vary reasoning but do not document energy savings.
Water disclosure
Water reporting has not been assessed for this choice.
Amp can route to different models and providers, so no single training dataset applies.
See the evidence
Select the exact model and provider before assessing training-data rights.
Amp documents multiple models and provider connections. A client-level rights grade would combine unrelated training datasets; prompt privacy controls don’t settle rights in model training data.
The selected model, version, provider, inherited base and post-training sources determine the rights assessment.
Amp runs a local client, while prompts, selected code context, thread records and tool results pass through its cloud service.
See the evidence
Amp’s security reference says the Amp server stores thread data and sends requests to listed model providers. Minimal retention is available for eligible managed routes; linked personal subscriptions use separate provider controls. Amp does not offer self-hosting.
Provider retention varies by route. Workspace admins can access private threads under audited controls, and BYOK still passes through Amp’s server.
Amp offers a local CLI and apps connected to a hosted agent service; no self-hosted service is documented.
See the evidence
The product docs describe a local client and managed server. Model selection is provided through third-party providers, but the reviewed sources do not describe a self-hosted server or open service license.
The full current product license and permission to run the server independently are not established.
Amp has no workforce disclosures covering the AI supply chain in the reviewed public sources.
See the evidence
The reviewed public sources do not describe employee conditions, worker voice, or contractor protections for Amp. Missing disclosure is an evidence gap, not proof of mistreatment.
1 + 0 + 0 + 0 = 1/10. Documented worker protections; missing disclosure is an evidence gap.
Contractor conditions, data annotation and moderation, worker representation, remedies and independent review remain unknown.
- Pay & conditions0/3
- No worker conditions policy or implementation evidence was found in reviewed sources.
- Worker voice & remedy0/3
- No worker voice, representation or grievance safeguard was found in reviewed sources.
- Evidence & follow-through0/3
- No independent worker-level evidence was found in reviewed sources.
No military relationship was established from the reviewed public sources for Amp; the score reflects uncertainty.
See the evidence
The sources reviewed did not establish a military contract or operational defense offering for Amp. The method starts at 7 when a military relationship is not established; missing disclosure is not proof that none exists.
7 base + 0 safeguard adjustment = 7/10 (0 of 9 safeguard points; every 3 adds 1). No military relationship was established in the cited sources. This ranks above known defense ties without claiming a verified clean record.
This bounded review cannot rule out unreported customers, indirect use, subcontracting or future work.
- Use restrictions0/4
- No public military-use restriction was found in the reviewed sources.
- Enforcement & oversight0/3
- No independently evidenced military-use oversight was found.
- Relationship disclosure0/2
- No military customer or relationship disclosure was identified.
No military relationship for Amp was established in the reviewed sources; this is not proof that none exists.
Military rubric ↗Amp was not named on the two BNC campaign pages checked. This limited list check is not an endorsement.
See the evidence
The BNC selects specific company targets; Amp was not named in the official priority guide or No Tech campaign when checked. This is not a nationality-based assessment or proof that the company has no other relevant relationships.
This covers only the BNC corporate guide and No Tech page. Other BDS campaigns, local calls, ownership changes, contractors and business relationships are outside the check.