# Runpod

The AI Developer Cloud

Source: https://wheredoirunthis.com/products/runpod · JSON: https://wheredoirunthis.com/api/products/runpod · Read on 2026-10-15

## At a glance

- Type: GPU cloud
- Kind: Hosted
- Isolation: Container
- Free tier: No
- Starting price: Not disclosed
- GPU: Yes
- Regions: 16
- Coverage: 21 of 61 assessed capabilities supported
- Agent readiness: read 11 of 11 signals, drive 11 of 11 signals

## Links

- Homepage: https://www.runpod.io/
- Docs: https://docs.runpod.io/
- Pricing: https://www.runpod.io/pricing
- Repository: https://github.com/runpod

## Pricing

- Model: Usage-based prepaid credits; enterprise post-paid contract option (beta); per-second compute, separate storage and request charges.
- Unit: GPU-hour (per GPU type, billed per second); storage GB-month; Public Endpoints per token / per request / per character / per megapixel / per second
- Free tier: No

| Tier | Price | Includes | Notes |
| --- | --- | --- | --- |
| Pods — On-demand GPU (Secure Cloud) | Secure Cloud listed examples: RTX A5000 $0.27/hr, RTX 4090 $0.89/hr, H100 PCIe $2.89/hr, H100 SXM $3.99/hr, H200 $5.29/hr, B200 $7.99/hr, B300 $8.99/hr | GPU container with optional persistent storage; SSH and HTTP/TCP connectivity; storage billed separately.; Pure usage; no base fee. Storage billed separately. Savings plans available for committed/extended usage. |  [source](https://www.runpod.io/pricing) |
| Serverless (flex/active workers) | Serverless listed flex examples: 4090 $1.10/hr, A100 $2.72/hr, RTX 6000 Pro $3.49/hr, H100 $4.79/hr, H200 $5.93/hr, B200 $8.64/hr, B300 $9.98/hr | Autoscaling GPU endpoint; zero active workers allows scale-to-zero, with storage billed independently.; No base fee; worker execution and initialization bill by second. Zero compute idle cost only when no active workers. |  [source](https://www.runpod.io/pricing) |
| Instant Clusters | Multi-node, up to 64 GPUs, per-second: H200 SXM $4.31/hr, A100 SXM $1.79/hr; L40S / H100 SXM / B200 contact-sales | Launch multi-GPU clusters in minutes, no commitment; attach shared storage; pay only for what you use.; Per-GPU-hour, multi-node. Larger / guaranteed capacity -> Reserved Clusters. |  [source](https://www.runpod.io/pricing) |
| Reserved Clusters / Enterprise | Contact sales — 1/3/6/12mo+ commitments, scale to 10,000+ GPUs, SLA-backed uptime, discounted rates | Dedicated clusters, guaranteed availability, custom configs, enterprise discounts, business invoicing.; Quote-based; not publicly priced. |  [source](https://www.runpod.io/pricing) |

- Overages: Pods listed $0.27–$8.99/GPU-hour; Serverless listed $0.58–$9.98/GPU-hour. Storage and beta global-volume request charges separate; no data transfer fees.
- Card required: No
- Source: https://docs.runpod.io/accounts-billing/billing.md

## Website design

Desktop public website · assessed 2026-10-15

Complete assessment.

Dark purple marketing pages combine luminous calls to action and a platform diagram with spacious centered pricing copy. Documentation switches to a white, structured sidebar-and-card reading layout. All three desktop roles captured; cookie overlays remain visible.

- Styles: Corporate, Technical illustration
- Landing page: captured
- Pricing: captured
- Docs: captured

## Runtime

- **GPU & accelerators**: Yes — "Current pricing includes B300, B200, H200, H100, A100, RTX Pro 6000 and consumer GPUs." [source](https://www.runpod.io/pricing) (confidence: high)
- **BYOC (Bring your own cloud)**: partial — "Separate limited-access Hybrid Cloud offering runs host agent on owned hardware or cloud VMs; hosted orchestration, enterprise annual management fee. Primary public GPU-cloud remains hosted." [source](https://www.runpod.io/hybrid-cloud) (confidence: high)
- **Secrets management**: Runpod Secrets — securely store and manage API keys, passwords, and tokens; injected into Pods/endpoints as environment variables — "Docs: 'Securely store and manage sensitive information like API keys, passwords, and tokens with Runpod secrets.'" [source](https://docs.runpod.io/) (confidence: high)
- **Storage options**: Container/volume disk $0.10/GB-month running; stopped volume disk $0.20; standard network volumes $0.07/GB-month below 1TB and $0.05 above; high-performance $0.14. Beta global volumes also bill Class A $0.005/1000 and Class B $0.0005/1000 requests. [source](https://www.runpod.io/pricing)
- **Supported workloads**: ai-inference, ai-training, fine-tuning, distributed-training, batch-jobs, model-serving — "Pods (dev/training/fine-tuning/batch), Serverless (containerized inference behind an API), Clusters (distributed training / multi-host inference), Public Endpoints (pre-deployed model serving)." [source](https://www.runpod.io/) (confidence: high)
- **Persistence**: Volume disk survives stop/restart but not Pod deletion; network volumes are independently persistent. Beta global volumes persist across Pod/worker deletion and are region-independent; object-backed, limited POSIX semantics. [source](https://docs.runpod.io/storage/globalvolume/overview.md)
- **Data jurisdiction**: 31 global regions; customer-selectable placement; compliance varies by data center. [source](https://www.runpod.io/)
- **Resource limits & quotas**: Default account spend cap $80/hour; new Pod needs at least one hour of selected configuration credits. Serverless operation-specific rate limits; job TTL and execution timeout max 7 days. Network volumes can be deleted on unpaid balance; global volume deletion after 15 days at $0. Cluster capacities differ by offering; enterprise page advertises 200+ on-demand / 10,000 reserved while pricing table still says 64. [source](https://docs.runpod.io/accounts-billing/billing.md)
- **SDK languages**: python, javascript, go — "Official Serverless SDK documentation explicitly lists Python, JavaScript and Go." [source](https://docs.runpod.io/serverless/sdks.md) (confidence: high)
- **Network & egress controls**: No — "No configurable outbound firewall / egress allowlist documented (sandbox-style). NOTE: ingress/egress data transfer is FREE (a pricing positive), which is distinct from egress CONTROL." [source](https://docs.runpod.io/) (confidence: medium)
- **Snapshots & forks**: No — "No running-instance snapshot/fork documented; image templates and build rollback are not instance snapshots." [source](https://docs.runpod.io/llms.txt) (confidence: medium)
- **Deployment methods**: container-image, dockerfile, git-push, cli, sdk, api, mcp, dashboard, template — "GitHub deployment builds a Dockerfile; updates trigger on new releases, not every push. Other surfaces documented in index, CLI and MCP docs." [source](https://docs.runpod.io/serverless/workers/github-integration.md) (confidence: high)
- **Isolation**: Container — "Compliance page: 'Our Secure Cloud and Serverless offerings only support Docker containers.' Pods/Serverless run user Docker containers (no gVisor/Kata/Firecracker disclosed). Bare Metal allows VMs/k8s as a SEPARATE offering; MIG partitions a GPU but is not a code-isolation boundary." [source](https://www.runpod.io/legal/compliance) (confidence: high)

## Capabilities

### AI & agents

| Capability | Status | Evidence |
| --- | --- | --- |
| Agent skills | Supported | "Install the skills with npx skills add runpod/runpod-plugins-official; router and focused GPU-management skills with native Claude Code/Codex/Gemini integration." [source](https://docs.runpod.io/get-started/agent-skills.md) |
| AI app builder | Not supported | "No prompt-to-app / vibe-coding builder." [source](https://www.runpod.io/) |
| Autonomous QA agent | Not supported | "No first-party autonomous QA agent." [source](https://www.runpod.io/) |
| Built-in coding agent | Not supported | "No first-party coding agent (it integrates external agents via MCP)." [source](https://www.runpod.io/) |
| BYOK for LLMs | Not supported | "Not a gateway product; no bring-your-own LLM-provider-key feature." [source](https://www.runpod.io/) |
| Credential-injection proxy | Not supported | "Not documented in fresh full documentation export and index; secrets are injected into workloads, not swapped outside them on the wire." [source](https://docs.runpod.io/llms.txt) |
| LLM gateway | Not supported | "Public Endpoints exposes a provider-hosted model catalog including several model vendors; no evidence of routing customer requests using external-provider credentials." [source](https://www.runpod.io/pricing) |
| MCP client | Not supported | "RunPod exposes MCP servers; it is not itself an MCP client." [source](https://docs.runpod.io/overview) |
| MCP-server hosting | Not supported | "Any container (incl. an MCP server) can run as a Pod/Serverless endpoint, but there is no dedicated MCP-server-hosting feature." [source](https://docs.runpod.io/overview) |
| MCP server | Supported | "Hosted infrastructure MCP at https://mcp.getrunpod.io/ with OAuth or API key; local @runpod/mcp-server; separate docs MCP." [source](https://docs.runpod.io/get-started/mcp-servers.md) |
| Model inference API (first-party) | Supported | "Public Endpoints: pre-deployed models accessible by API without infrastructure setup; model-specific token/request pricing." [source](https://www.runpod.io/pricing) |
| Preinstalled coding agent | Not supported | "Not documented in fresh full documentation export and index; plugins manage infrastructure from external agents; no platform-preinstalled Pod harness evidenced." [source](https://docs.runpod.io/llms.txt) |

### Networking & delivery

| Capability | Status | Evidence |
| --- | --- | --- |
| Automatic HTTPS | Supported | "Pod HTTP proxy provides automatic HTTPS endpoints (https://{podid}-{port}.proxy.runpod.net)." [source](https://docs.runpod.io/overview) |
| Custom domains | Not supported | "No BYO custom domain for Pod/endpoint URLs documented (proxy URLs only)." [source](https://docs.runpod.io/overview) |
| Direct public IPv6 | Not supported | "TCP-port mapping via proxy + optional public IP (stable for Secure Cloud Pods, IPv4); no stable per-instance public IPv6 for raw TCP/UDP." [source](https://docs.runpod.io/overview) |
| Edge KV store | Not applicable | "Not an edge runtime -> edge-layer key not applicable." [source](https://www.runpod.io/) |
| Egress controls | Not supported | "No configurable workload outbound allowlist/firewall found in fresh docs export. Private networking and free transfer do not establish egress policy." [source](https://docs.runpod.io/llms.txt) |
| Email send / receive | Not supported | "No email primitive." [source](https://www.runpod.io/) |
| Public preview URLs | Partially supported | "Each Pod gets a shareable public proxy URL to its running HTTP service; not per-deploy/branch preview semantics." [source](https://docs.runpod.io/overview) |
| Private networking | Supported | "Account-private Pod network across 17 documented data centers; private IP and POD_ID.runpod.internal DNS; NVIDIA GPU Pods only, 100 Mbps." [source](https://docs.runpod.io/pods/networking.md) |
| Built-in reverse proxy | Supported | "RunPod proxy routes external traffic to a Pod's exposed HTTP/TCP ports." [source](https://docs.runpod.io/overview) |

### Organisation & security

| Capability | Status | Evidence |
| --- | --- | --- |
| Built-in auth | Partially supported | "Organization resource ownership, role-based permissions and platform access control; not a managed end-user authentication service for applications." [source](https://docs.runpod.io/organizations/orgs-overview.md) |
| BYOC | Partially supported | "Limited-access Hybrid Cloud uses an outbound-only host agent on owned hardware or VMs in your own cloud account; managed orchestration and annual management fee. Separate offering from primary rented GPU compute." [source](https://www.runpod.io/hybrid-cloud) |
| Deploy into customers' VPCs | Not supported | "Not documented in fresh full documentation export and index; Hybrid Cloud manages your own compute, not a SaaS vendor deploying into its customers accounts." [source](https://docs.runpod.io/llms.txt) |
| Secrets management | Supported | "Runpod Secrets — 'Securely store and manage sensitive information like API keys, passwords, and tokens.'" [source](https://docs.runpod.io/overview) |
| Self-host / OSS core | Not supported | "Hosted; only SDKs/CLI/providers/templates are OSS — no self-hostable cloud core (-> profile_kind hosted)." [source](https://github.com/runpod) |
| Sharing | Partially supported | "Share a running Pod via its public proxy URL; publish reusable templates to the Runpod Hub (creator program)." [source](https://www.runpod.io/product/runpod-hub) |
| SSO / SAML | Partially supported | "Enterprise SSO/SAML with Okta, Entra ID, Auth0, Google Workspace or Duo; our team sets it up with you directly." [source](https://www.runpod.io/enterprise) |
| Teams | Supported | "Teams with member roles/permissions + cost centers for per-team/project billing attribution." [source](https://docs.runpod.io/accounts-billing/manage-accounts) |
| Workload identity | Not supported | "Not documented in fresh full documentation export and index; API keys and stored secrets documented, not workload-scoped federated short-lived identity." [source](https://docs.runpod.io/llms.txt) |

### Developer access

| Capability | Status | Evidence |
| --- | --- | --- |
| CLI | Supported | "Official runpodctl installation via cli.runpod.net or Homebrew; resource management and transfer tools." [source](https://docs.runpod.io/get-started/agent-skills.md) |
| E2B-compatible API | Not supported | "Not documented in fresh full documentation export and index; Runpod APIs are native, not advertised E2B-compatible." [source](https://docs.runpod.io/llms.txt) |
| Filesystem API | Partially supported | "S3-compatible API provides remote network-volume file access without a Pod; not arbitrary running-container filesystem access." [source](https://docs.runpod.io/storage/s3-api.md) |
| Git deploy | Supported | "Import Git Repository, choose branch and Dockerfile; Runpod builds and deploys automatically. Subsequent updates require a new GitHub release, not every push." [source](https://docs.runpod.io/serverless/workers/github-integration.md) |
| HTTP data API | Not applicable | "No first-party stored-data plane -> data-layer key not applicable." [source](https://www.runpod.io/) |
| OpenAPI spec | Supported | "Fetchable OpenAPI 3.0.3 spec (154KB) — 'Public Rest API for managing Runpod programmatically.'" [source](https://docs.runpod.io/api-reference/openapi.json) |
| Private container registry | Not supported | "Supports container-registry AUTH to PULL images from external private registries; does not HOST a private registry." [source](https://docs.runpod.io/api-reference/container-registry-auths/GET/containerregistryauth) |
| Public API | Supported | "Public REST API (rest.runpod.io/v1) + GraphQL — manage Pods/endpoints/templates/volumes/registries." [source](https://docs.runpod.io/api-reference/overview) |
| SDK | Supported | "Official Serverless SDKs documented for Python, JavaScript and Go." [source](https://docs.runpod.io/serverless/sdks.md) |
| SSH access | Supported | "SSH into Pods with secure key generation; SSH Configuration Guide; expose port 22." [source](https://docs.runpod.io/pods/configuration/use-ssh) |

### Compute & storage

| Capability | Status | Evidence |
| --- | --- | --- |
| Cloud-init / ignition config | Not supported | "Container-based — launch customization is via Docker image/template + environment variables, not cloud-init/ignition user-data." [source](https://docs.runpod.io/overview) |
| Desktop / computer-use sandbox | Not supported | "No first-party computer-use/virtual-desktop sandbox (some community templates ship noVNC GUIs, but not a product feature)." [source](https://www.runpod.io/) |
| GPU compute | Supported | "GPU pricing lists B300, B200, H200, H100, A100 and RTX GPU options." [source](https://www.runpod.io/pricing) |
| Live-state (memory) snapshots | Not supported | "No public documentation of restoring Pod/worker memory and running process checkpoints; FlashBoot pre-warmed workers do not establish memory snapshots." [source](https://docs.runpod.io/llms.txt) |
| Multi-region | Supported | "'31 global regions' / '30+ regions'; customer-selectable data-center location." [source](https://www.runpod.io/) |
| Object storage | Supported | "Access network-volume files through S3-compatible API without launching a Pod; separate S3 keys. Beta global volumes are object-backed mounted storage, not full POSIX." [source](https://docs.runpod.io/storage/s3-api.md) |
| Persistent disk | Supported | "Network Volumes (Standard/High-Performance) + Volume Disk; /workspace persistent dir." [source](https://www.runpod.io/pricing) |
| Sandbox templates / custom images | Supported | "Reusable Pod templates bundle prebuilt container images, hardware specifications and networking settings." [source](https://docs.runpod.io/pods/templates/overview.md) |
| Scale to zero | Supported | "Serverless flex workers=(0,n) scale to zero — 'Zero idle cost. Your endpoint costs nothing when it's not running.' FlashBoot sub-200ms cold start." [source](https://www.runpod.io/product/serverless) |
| Selectable isolation runtime | Not supported | "Not documented in fresh full documentation export and index; container runtime documented; no per-workload choice of isolation boundary." [source](https://docs.runpod.io/llms.txt) |
| Snapshots / forking | Not supported | "No snapshot / copy-on-write fork of a running Pod; reuse is image/template-based." [source](https://docs.runpod.io/overview) |

### Data

| Capability | Status | Evidence |
| --- | --- | --- |
| Database branching | Not applicable | "No managed database -> data-layer key not applicable." [source](https://www.runpod.io/) |
| Ephemeral instant DB (no-signup) | Not applicable | "No database -> data-layer key not applicable." [source](https://www.runpod.io/) |
| First-party database | Not supported | "No managed database service (compute + object/network storage only)." [source](https://www.runpod.io/) |
| Vector search | Not applicable | "No first-party data layer / vector store -> data-layer key not applicable (you can run a vector DB in a Pod yourself)." [source](https://www.runpod.io/) |

### Platform & operations

| Capability | Status | Evidence |
| --- | --- | --- |
| Managed CI/CD runners | Not supported | "No drop-in GitHub Actions / GitLab CI runner offering." [source](https://www.runpod.io/) |
| Built-in observability | Supported | "Built-in Grafana dashboards for GPU, CPU, memory, network and storage; Serverless logs and metrics also documented." [source](https://docs.runpod.io/instant-clusters/cluster-observability.md) |
| Preview environments | Not supported | "Not a per-PR/branch ephemeral-environment platform." [source](https://www.runpod.io/) |
| Scheduled tasks | Not supported | "No first-party cron/scheduler primitive (Serverless is request/queue-triggered)." [source](https://docs.runpod.io/overview) |
| Project template marketplace | Not supported | "Runpod Hub is a COMPUTE template/model/OSS-app catalogue (maps to sandbox-templates), not a no-code project-STARTER marketplace (the Framer/Base44 sense)." [source](https://www.runpod.io/product/runpod-hub) |
| Durable workflows | Not supported | "No durable-execution/workflow engine (Serverless queues tasks but is not durable workflows)." [source](https://www.runpod.io/) |

## Agent readiness

### Read: 11 of 11 signals present

11/11 signals (7 full, 4 partial)

- Llms txt: pass — "Substantive apex llms.txt and 82KB docs llms.txt. Source: https://www.runpod.io/llms.txt; confidence: high."
- Llms full txt: pass — "Comprehensive 1,817,476-byte docs export retained as bounded POINTER. Source: https://docs.runpod.io/llms-full.txt; confidence: high."
- Robots AI policy: partial — "User-agent: * Allow: / # Content Signals (https://contentsignals.org/): # allow search and AI assistant use of this content, disallow AI training. Content-Signal: search=yes, ai-input=yes, ai-train=no # Pagination variants (*_page params) are crawlable ON PURPOSE: # a head script serves noindex,follow for them so Google deindexes # the leaked variants. Do not re-add Disallow rules for *_page until # GSC "Indexed, though blocked by robots.txt" reaches 0. # LLM discovery file # For language model indexing, see: # https://www.runpod.io/llms.txt Sitemap: https://www.runpod.io/sitemap.xml Source: https://www.runpod.io/robots.txt; confidence: high."
- JSON LD: partial — "WebPage graph, ImageObject and breadcrumbs; no Product with offers. Source: https://www.runpod.io/; confidence: high."
- Sitemap: pass — "215,856-byte sitemap with many URLs; bounded POINTER retained. Source: https://www.runpod.io/sitemap.xml; confidence: high."
- Meta quality: pass — "Descriptive title, description and OpenGraph tags in homepage HTML. Source: https://www.runpod.io/; confidence: high."
- Markdown negotiation: pass — "Accept: text/markdown returns clean markdown on docs API overview; apex still HTML. Source: https://docs.runpod.io/api-reference-v2/overview.md; confidence: high."
- Content efficiency: partial — "Readable text bytes / raw HTML bytes = 11930/309251 = 3.858%. Source: https://www.runpod.io/; confidence: high."
- Docs discoverable: pass — "Docs is a first-class header navigation link. Source: https://www.runpod.io/; confidence: high."
- Pricing machine readable: pass — "Structured GPU, Serverless, cluster and storage pricing tables; updated October 8, 2026. Source: https://www.runpod.io/pricing; confidence: high."
- Anti bot: partial — "Ordinary public fetches and rendered retries succeeded; apex Accept:text/markdown request returned HTTP 403, then foxcrawl retry succeeded. Intermittent/request-specific restriction. Source: https://www.runpod.io/; confidence: high."

### Drive: 11 of 11 signals present

11/11 signals (7 full, 4 partial)

- Public API: pass — "REST API v2 documented; v1 retires November 15, 2026. Source: https://docs.runpod.io/api-reference-v2/overview.md; confidence: high."
- Openapi spec: pass — "Fetchable valid OpenAPI v2 spec; 274,963 bytes retained POINTER. Source: https://api.runpod.io/v2/openapi.json; confidence: high."
- MCP server: pass — "Hosted https://mcp.getrunpod.io/ and installable @runpod/mcp-server; docs MCP separately. Source: https://docs.runpod.io/get-started/mcp-servers.md; confidence: high."
- SDK packages: pass — "Official Python, JavaScript and Go SDKs. Source: https://docs.runpod.io/serverless/sdks.md; confidence: high."
- Auth docs: pass — "API keys, scopes, SSH and S3 credentials explicitly documented. Source: https://docs.runpod.io/get-started/credentials.md; confidence: high."
- Agent self serve creds: partial — "Dashboard-issued API keys; hosted MCP OAuth requires browser sign-in/approval; no unattended issuance documented. Source: https://docs.runpod.io/get-started/mcp-servers.md; confidence: high."
- Webhooks: partial — "Webhook job completion callback, 200 acknowledgment and retries documented; signing not evidenced. Source: https://docs.runpod.io/serverless/endpoints/send-requests.md; confidence: high."
- CLI tool: pass — "Official runpodctl CLI documented with installation and resource management commands. Source: https://docs.runpod.io/get-started/agent-skills.md; confidence: high."
- Error model: pass — "REST v2 errors use RFC 9457 title, status, detail and optional validation errors array. Source: https://docs.runpod.io/api-reference-v2/migrate-from-v1.md; confidence: high."
- Rate limit docs: partial — "Operation-specific limits and dynamic worker-based limits; 429 documented, response headers not established. Source: https://docs.runpod.io/serverless/endpoints/send-requests.md; confidence: high."
- Agent config: partial — "Installable official skills/plugin, plus one-shot agent setup; no multiple distinct standard discovery surfaces proven. Source: https://docs.runpod.io/agent-setup.md; confidence: high."

## Coding harnesses

- Claude Code (Anthropic) — "Supported clients: Claude Code, Codex CLI, Cursor, VS Code with Copilot, Windsurf, Cline, Gemini CLI." [source](https://docs.runpod.io/get-started/mcp-servers.md)
- Codex CLI (OpenAI) — "Supported clients: Claude Code, Codex CLI, Cursor, VS Code with Copilot, Windsurf, Cline, Gemini CLI." [source](https://docs.runpod.io/get-started/mcp-servers.md)
- Cursor (Anysphere) — "Supported clients: Claude Code, Codex CLI, Cursor, VS Code with Copilot, Windsurf, Cline, Gemini CLI." [source](https://docs.runpod.io/get-started/mcp-servers.md)
- GitHub Copilot (GitHub) — "Supported clients: Claude Code, Codex CLI, Cursor, VS Code with Copilot, Windsurf, Cline, Gemini CLI." [source](https://docs.runpod.io/get-started/mcp-servers.md)
- Windsurf (Codeium) — "Supported clients: Claude Code, Codex CLI, Cursor, VS Code with Copilot, Windsurf, Cline, Gemini CLI." [source](https://docs.runpod.io/get-started/mcp-servers.md)
- Cline (Cline) — "Supported clients: Claude Code, Codex CLI, Cursor, VS Code with Copilot, Windsurf, Cline, Gemini CLI." [source](https://docs.runpod.io/get-started/mcp-servers.md)
- Gemini CLI (Google) — "Supported clients: Claude Code, Codex CLI, Cursor, VS Code with Copilot, Windsurf, Cline, Gemini CLI." [source](https://docs.runpod.io/get-started/mcp-servers.md)
- OpenCode (opencode) — "Configure OpenCode to use Runpod Public Endpoints as the model provider." [source](https://docs.runpod.io/public-endpoints/ai-coding-tools.md)

Plausible compatibility (derived, not stored):

- Aider
- Amp
- Antigravity
- Devin
- Droid
- Goose
- Hermes Agent
- Kilo Code
- Kiro CLI
- Pi
- Qwen Code

## Compliance & regions

- SOC 2 Type II: certified (scope: hosted) — "Runpod holds SOC 2 Type 2" [source](https://trust.runpod.io/)
- SOC 3: certified (scope: hosted) — "Runpod holds SOC 3" [source](https://trust.runpod.io/)
- ISO/IEC 27001: certified (scope: hosted) — "Runpod holds ISO/IEC 27001:2022" [source](https://trust.runpod.io/)
- HIPAA: self-attested (scope: hosted) — "maintains active HIPAA and GDPR programs" [source](https://trust.runpod.io/)
- GDPR: self-attested (scope: hosted) — "maintains active HIPAA and GDPR programs" [source](https://trust.runpod.io/)
- NIS 2: self-attested (scope: hosted) — "Documents: NIS2 Management Attestation Letter" [source](https://trust.runpod.io/)
- Regions: EU-CZ-1 (Czech Republic), EU-FR-1 (France), EU-NL-1 (Netherlands), EU-RO-1 (Romania), EU-SE-1 (Sweden), EUR-IS-2 (Iceland), OC-AU-1 (Australia), US-CA-2 (California), US-GA-1 (Georgia), US-GA-2 (Georgia), US-IL-1 (Illinois), US-KS-2 (Kansas), US-NC-1 (North Carolina), US-TX-3 (Texas), US-TX-4 (Texas), US-WA-1 (Washington)
- Regions note: Homepage advertises 31 regions. Listed entries are the documented 17-data-center PRIVATE NETWORKING subset, not an exhaustive compute region inventory. Iceland is not an EU member.

## Company & presence

- Headquarters: 329 Bryant St #4D, San Francisco, CA 94107 (published address; headquarters designation not explicit)
- Founded: 2022 — co-founders Zhen Lu (CEO) and Pardeep Singh (CTO), ex-Comcast developers
- Funding: Seed, $20M, 2024-05
- Funding: Growth investment led by Summit Partners, $100M, 2026-06-24
- Privacy policy: https://www.runpod.io/legal/privacy-policy
- Terms of service: https://www.runpod.io/legal/terms-of-service
- GitHub: https://github.com/runpod
- Docs: https://docs.runpod.io/
- Blog: https://www.runpod.io/blog
- Discord: https://discord.gg/pJ3P2DbUUq

## Nearby in the atlas

- [CoreWeave (CoreWeave Sandboxes)](/products/coreweave.md)
- [Nebius](/products/nebius.md)
- [AgentNest](/products/agentnest.md)
- [Anthropic Sandbox Runtime (srt)](/products/sandbox-runtime.md)
- [Aptible](/products/aptible.md)
- [Canine](/products/canine.md)

## Method

Facts come from the product's own public pages. Each claim carries its evidence: a quote and the page it came from. We count capabilities rather than grade products, and "Not assessed" means we have not checked, never that the product lacks it.