NVCF (NVIDIA Cloud Functions)

Open source toolServerless

Self-managed GPU functions, streaming and batch workloads on Kubernetes

Homepage Docs Repository

Scanned 15 Oct 2026 · Schema v5 · features v21

Licence: Apache-2.0
Self-hosted; you pay for the infrastructure.
Free tier: Free to run
Isolation: Container
shares the host kernel
Regions: —
Not disclosed on the public surface.
Capabilities: 14 / 61
12 partial · 35 fail · 0 not assessed
Agent readiness: 9/10 · 9/10
read · drive, pre-login surface

Advertised for

GPU inference · GPU streaming · Batch GPU jobs · Fine tuning · Evaluation · Data preparation

Built for

Developers, then technical

· Container/Helm packaging and CLI/API lifecycle management.
· Kubernetes, GPU operators, credentials and multi-cluster infrastructure setup.

01 Website design Desktop public website · assessed 15 Oct 2026

Complete assessment.

The public entry point is GitHub's dense repository interface, while NVIDIA's docs use a white, three-column technical layout with restrained green branding, clear heading hierarchy and a prominent NVCF banner. Pricing is not applicable to this self-host-only OSS subject.

Corporate

Landing page: captured
Pricing: not applicable
Docs: captured

02 Runtime Each fact carries its source; dots show confidence. Open a row for the evidence.

Isolation & lifecycle

Isolation primitive

Containershares the host kernel

Persistence

Operator-managed persistent infrastructure; Helm functions can request PVC-backed volumes. No universal per-function durable-disk lifecycle guarantee.

Snapshots / forks

Optional beta NvSnap GPU checkpoint/restore; single-GPU full process restore, any-GPU-count warm-cache restore

Resources & limits

Storage

Platform state in Cassandra; OpenBao Raft persistence; shared read-only model cache through SMB/CSI/RWX; Helm functions may use PVCs; self-hosted task result upload is unsupported.

Limits & quotas

Per-function-version autoscaling with minInstances>=0 and maxInstances>0; default idle timeout 1800s (operator-configurable); API-key scopes and account quotas are administered locally.

GPU

Yes

Access

Deploy interface

container-imagecliapi

SDK languages

PythonGo

Network & secrets

Egress controls

NVCA-managed customizable Kubernetes NetworkPolicy: default public-internet egress with common private ranges blocked; CNI enforcement required.

Secrets

ESS tenant-scoped encrypted secrets plus ESS Agent workload injection; OpenBao wraps namespace keys with AES-256-GCM master keys.

Bring your own cloud

Placement & use

Jurisdiction

Advertised workloads

GPU inferenceGPU streamingBatch GPU jobsFine tuningEvaluationData preparation

03 Capabilities 14 pass · 12 partial · 35 fail · 61 of 61 assessed

Supported — Documented on the public surface and verified.
Partially supported — Exists with caveats: gated, thin, or enterprise-only.
Not supported — Checked and not offered.
Not assessed — Scanned before this capability was tracked. Not a lack.

Compute & storage 3/11

Persistent diskStorage that survives restart/stop (not ephemeral-only).

Partially supported

Object storageManaged S3-compatible / blob object store the platform PROVIDES to a user's workloads (buckets, presigned URLs). DISTINCT from persistent-disk (block/volume attached to an instance) and from a self-host dependency (source.self_host.deps[]=["object-storage"] is an infra dep to RUN an OSS tool — the opposite direction). SCOPE varies and matters: native first-class (Railway Buckets, Cloudflare R2) vs a sibling SKU / extension (Fly Tigris, Bunny Storage, DigitalOcean Spaces) — record which in the evidence quote / note. Admitted v15.

Not supported

Snapshots / forkingSnapshot or copy-on-write fork of a running instance (filesystem-only counts). Whether the snapshot also restores memory + running processes is recorded separately by memory-snapshot (v19).

Partially supported

Live-state (memory) snapshotsA snapshot / checkpoint / pause captures a running instance's MEMORY and PROCESS state and restores it, so running processes continue where they left off instead of a freshly booted instance with the same files. Counts whether user-triggered (snapshot, checkpoint, pause-to-storage) or used to restore pre-initialised state (memory snapshots for fast cold starts). Distinct from snapshots-forks (any snapshot or fork, including filesystem-only; a product passing that key may still fail this one) and from scale-to-zero (an idle-cost property; suspension may or may not keep memory). Actor multiplexing (see scale-to-zero) records its live-state half here — only when the suspended actor's MEMORY/process state is what comes back, not a freshly booted worker with the actor's files.

Partially supported

Scale to zeroIdle instances suspend at no / negligible cost. ACTOR MULTIPLEXING (many suspended stateful actors sharing a small pool of pre-warmed workers, woken on request — Agent Substrate, AX, kagent; Rivet's hibernating actors) has NO key of its own (proposal REJECTED v20) — score the idle-cost half HERE and the live-state half under memory-snapshot; the oversubscription ratio / density is a performance characteristic the registry does not key.

Supported

GPU computeGPU-backed instances available.

Supported

Multi-regionCompute placeable in more than one geographic region.

Supported

Sandbox templates / custom imagesBuild reusable sandbox/VM images from an OCI image or Dockerfile (a base you fork instances from).

Not supported

Cloud-init / ignition configCustomise a VM / microVM / instance at boot via cloud-init (or ignition) user-data — the same boot-metadata interface as AWS/Azure/GCP. Inject packages, users, SSH keys, files, and run-commands declaratively at launch. Distinct from sandbox-templates (a PREBUILT image you fork instances from) — this is PER-LAUNCH declarative customisation, not a baked image.

Not supported

Desktop / computer-use sandboxA virtual desktop (GUI) for LLM computer-use — browser + desktop automation, often via noVNC/VNC.

Not supported

Selectable isolation runtimeThe user chooses the isolation runtime per workload / environment / sandbox from more than one option (e.g. microVM vs gVisor vs plain container, native namespaces vs gVisor, gVisor vs VM). Records the CHOICE; the isolation_primitive facet still records the best-evidenced primary. A platform that switches runtime automatically without a user choice (e.g. gVisor forced for GPU) does not pass on that alone.

Partially supported

Networking & delivery 2/9

Automatic HTTPSTLS endpoints provisioned automatically per workload.

Partially supported

Built-in reverse proxyPlatform routes external traffic to workload services.

Supported

Custom domainsBring-your-own domain via CNAME/DNS.

Partially supported

Public preview URLsShareable per-deploy / per-branch URLs.

Not supported

Direct public IPv6Per-instance stable public IPv6 for raw TCP/UDP (beyond a managed HTTPS proxy) — e.g. mosh / raw sockets.

Not supported

Egress controlsConfigurable OUTBOUND network policy — egress firewall / allowlist / egress proxy on traffic LEAVING a workload. Scoped to OUTBOUND only (v15): private-network REACHABILITY between a user's own services is private-networking, not this. Static-egress-IP belongs here (an outbound concept).

Supported

Private networkingA platform-internal private network between a user's OWN services — service-to-service connectivity, VPC, service mesh, and/or service discovery, not exposed to the public internet. DISTINCT from egress-controls (OUTBOUND firewall/allowlist), reverse-proxy (PUBLIC inbound ingress), and direct-ipv6 (per-instance PUBLIC addressing). Admitted v15.

Partially supported

Email send / receiveSend and/or receive email from the workload.

Not supported

Edge KV storeA globally-replicated key-value store read from WITHIN an edge runtime at the PoP (low-latency, eventually-consistent) — Cloudflare Workers KV, Deno KV, Fastly KV Store, Akamai EdgeKV, Vercel Edge Config. Distinct from first-party-db (a managed DATABASE service you provision and connect to) and from http-query-api (HTTP access to an external store) — edge-kv is colocated with the edge compute and read at the edge.

Not supported

Developer access 3/10

Public APIDocumented programmatic API.

Supported

OpenAPI specMachine-readable OpenAPI / Swagger spec published.

Supported

CLIOfficial command-line tool.

Supported

SDKOfficial client library/libraries.

Partially supported

SSH accessDirect SSH into workloads.

Not supported

Filesystem APIRemote file operations over the API (list/read/write/watch/copy/upload) without SSH.

Not supported

Git deployDeploy from a Git repo / push-to-deploy.

Not supported

Private container registryHosted private image registry.

Not supported

E2B-compatible APIDrop-in compatibility with the E2B sandbox API/SDKs — existing E2B client code runs against this product by changing the endpoint/domain + key, with no business-code changes. E2B itself passes as the reference implementation. Distinct from public-api / sdk (the product has its own API — this key says it also speaks another vendor's) and from self-host-oss (a self-hosted E2B-compatible server is scored on both).

Not supported

HTTP data APIQuery / read / write the stored DATA over plain HTTP/REST (or GraphQL) WITHOUT the native protocol driver (Postgres wire / Redis RESP) — agent- and edge-friendly data access via fetch(). The DATA plane; distinct from public-api (a control-plane / management API). Auto-generated-from-schema REST/GraphQL (PostgREST, pg_graphql) is a flavour of this.

Not supported

AI & agents 2/12

Built-in coding agentShips a first-party AI coding agent in the product.

Not supported

Agent skillsInstallable / declarable agent skills or tools.

Supported

AI app builderPrompt-to-app / vibe-coding generation surface.

Not supported

MCP serverExposes a Model Context Protocol server for agents.

Supported

MCP clientActs as an MCP client / consumes MCP tools.

Not supported

MCP-server hostingRun/host MCP servers as a workload on the platform (distinct from exposing a first-party MCP control server — that is mcp-server).

Not supported

LLM gatewayRoutes LLM API calls through the platform (multi-provider).

Partially supported

BYOK for LLMsBring-your-own LLM provider keys.

Not supported

Model inference API (first-party)A FIRST-PARTY hosted LLM/model inference API — the provider serves its OWN catalogue of (usually open-weight) models behind a (typically OpenAI-compatible) endpoint, billed per token. Distinct from llm-gateway (a multi-PROVIDER proxy/router for YOUR calls to external providers like OpenAI/Anthropic) and from byok-llm (bring your own provider keys) — here the platform IS the inference provider, serving its own models.

Not supported

Autonomous QA agentA first-party AUTONOMOUS AI agent that QA-TESTS a running app / implementation — exercising it like a real user (clicking through flows, checking behaviour) and reporting bugs/defects (often back to a coding agent, or before a human review). Distinct from built-in-agent (a code-AUTHORING agent — this one VERIFIES the result, it does not write the feature; a product can ship both), from ai-guided-code-review (organises a diff for a HUMAN to review — not an autonomous tester), and from observability (runtime metrics/logs, not active testing).

Not supported

Credential-injection proxyThe platform holds real secrets OUTSIDE the workload and injects them on the wire (an egress / broker proxy swaps a placeholder or phantom token for the real credential on matching outbound requests), so code running in the sandbox — e.g. an untrusted agent — never sees or can exfiltrate the real key. Distinct from secrets-management (storage + injection INTO the workload as env vars/files — the workload holds the secret) and from egress-controls (outbound allow/deny POLICY; a product can have an allowlist without injecting credentials, and vice versa).

Not supported

Preinstalled coding agentThe platform installs and launches a THIRD-PARTY coding agent (a harnesses.yml entry — Claude Code, Codex, OpenCode, Pi ...) inside a workspace/sandbox it provisions, with the agent's auth wired in (your own account or API key), so you start an agent session rather than an empty box. Distinct from built-in-agent (a FIRST-PARTY agent the vendor ships) and from harnesses[] (which records that a harness CAN target the product, not that the product runs it for you).

Not supported

Data 0/4

First-party databasePlatform-provided managed database.

Not supported

Database branchingCopy-on-write branch of a DATABASE (schema + data) as a dev-lifecycle primitive — branch your data like code, usually instant via a metadata operation. Distinct from snapshots-forks (forks a running compute INSTANCE / VM state) and from preview-environments (a full per-PR environment that may USE a DB branch). For a data product a CoW branch maps HERE, not to snapshots-forks.

Not supported

Vector searchFirst-class storage, indexing, and similarity search of vector embeddings — a native vector type/index or a managed vector DB. The data-layer primitive for RAG / semantic retrieval.

Not supported

Ephemeral instant DB (no-signup)Provision a working, usable instance via a single command / HTTP call with NO account, signup, or card — typically time-boxed (expires unless claimed). A zero-onboarding path an agent can hit directly.

Not supported

Platform & operations 0/6

Scheduled tasksCron / scheduled job primitive.

Partially supported

Durable workflowsDurable execution / long-running workflow primitive.

Not supported

Preview environmentsA full EPHEMERAL environment spun up per PR / branch — compute + addons + (often) a DB branch — torn down on merge. More than preview-urls (which is just a shareable URL for one deploy).

Not supported

Built-in observabilityFirst-party metrics / logs / tracing / APM / profiling included in the platform (no third-party stack required to see what a workload is doing).

Partially supported

Managed CI/CD runnersPlatform-hosted, drop-in CI/CD runner VMs (GitHub Actions / GitLab), typically faster and cheaper than the CI vendor's own hosted runners, billed per runner-minute; isolated + ephemeral per job. Distinct from git-deploy (push-to-deploy of an APP) and scheduled-tasks (cron) — this is managed CI COMPUTE you point your existing pipelines at.

Not supported

Project template marketplaceA marketplace / gallery of prebuilt PROJECT STARTER templates (apps or websites) a user clones or launches a new project from and then customises — the "start from a template" surface of a builder / PaaS. Distinct from sandbox-templates (a COMPUTE base IMAGE you fork runtime INSTANCES from — Dockerfile / OCI), from app-builder (the prompt-to-app GENERATION surface — a product can have both), and from one-click OSS-app catalogues / helm-addon-marketplace (deploy a FINISHED app as-is, not a starter you build on).

Not supported

Organisation & security 4/9

Built-in authPlatform-provided authentication / IAM for workloads.

Supported

SSO / SAMLEnterprise single sign-on.

Not supported

Secrets managementFirst-class secrets storage / injection.

Supported

TeamsMulti-user team accounts.

Partially supported

SharingShare a running workload/resource via link (optionally gated).

Not supported

BYOCRuns inside the customer's own cloud account.

Supported

Deploy into customers' VPCsA SaaS vendor uses the platform to deploy and operate ITS software inside ITS CUSTOMERS' cloud accounts / VPCs (nested or "customer BYOC"), managed from the vendor's control plane. Distinct from byoc (your own workloads in your own cloud account).

Not supported

Workload identityThe platform gives each workload its own verifiable identity (OIDC token, SPIFFE SVID, cloud IAM federation) that it exchanges for short-lived credentials, so no long-lived static secret is stored or injected. Distinct from secrets-management (stores / injects static secrets) and credential-injection-proxy (swaps real secrets in on the wire).

Not supported

Self-host / OSS coreOpen-source core that can be self-hosted.

Supported

Proposed, not verified

Capabilities the scan saw but the registry does not track yet. A key is admitted only once two distinct products exhibit it. These are not scored.

GPU fleet / multi-cluster managementmulti-cluster-management
“One control plane registers/manages multiple GPU clusters across providers/on-prem; cluster lifecycle, GPU discovery and capacity management are broader than geographic placement (multi-region).” github.com
Gang schedulinggang-scheduling
“Optional KAI/Grove integration places all required replicas atomically for coordinated multi-node GPU workloads; distinct from generic GPU availability.” docs.nvidia.com
Topology-aware GPU placementtopology-aware-gpu-scheduling
“Topology-aware placement selects a common GPU/NVLink clique for multi-node workloads. Requires optional scheduler/topology prerequisites; not a new registry admission.” docs.nvidia.com

04 Agent readiness Pre-login surface · counts, not a grade

Can an agent read it?

parse-ability · docs, metadata, machine-readable surfaces

9/10

5 full · 4 partial · 90% of applicable checks

llms.txt
Supported
“docs-1.0.3-llms.txt is a substantive complete current page index, discovered via docs-llms.txt version selector.”
llms-full.txt
Not supported
“Both root and version /llms-full.txt return HTTP 200 after redirect to the corresponding llms.txt index, not aggregated full content. Retained docs-llms-full.txt and docs-1.0.3-llms-full.txt; do not count 200 as full corpus.”
robots.txt AI policy
Supported
“docs-robots.txt explicitly allows major AI bots and clean .md/llms paths; no blanket Disallow /.”
Sitemap
Supported
“docs-sitemap.xml returns substantive NVCF URL set (28713 bytes).”
Markdown negotiation
Supported
“Accept: text/markdown on /overview/overview redirects to /overview/overview.md; Content-Type text/markdown; full page 2172 bytes retained docs-markdown-negotiation.txt.”
Docs discoverable
Supported
“github-readme.md has top-level Docs link, installation/API links and repository documentation map.”
Machine-readable pricing
Not applicable
“Pure oss-tool with no hosted commercial price tiers; Apache-2.0 + operator infrastructure cost, not missing hosted pricing.”
JSON-LD
Partially supported
“Recovered rendered HTML contains BreadcrumbList and TechArticle JSON-LD, but no Product/SoftwareApplication schema.”
Meta tags
Partially supported
“Rendered docs HTML has description and OpenGraph/Twitter image/title metadata; sampled description is 12 whitespace-delimited words, below rubric 20-word threshold.”
Content efficiency
Partially supported
“Recovered rendered HTML text/HTML ratio measured approximately 1.45%; large Fern scripts/markup reduce content efficiency.”
No anti-bot wall
Partially supported
“Plain urllib HTML returned HTTP 403; foxcrawl HTML/browser and native Markdown succeed. Intermittent access friction, not an unresolved content gap.”

Can an agent drive it?

act-ability · APIs, SDKs, MCP, credentials, webhooks

9/10

6 full · 3 partial · 90% of applicable checks

Public API
Supported
“docs-overview-api.md documents API groups and concrete endpoints/scopes on operator gateway.”
OpenAPI spec
Supported
“openapi.json is valid JSON with openapi=3.1.0 and info.version=2.261.2, 116822 bytes.”
MCP server
Supported
“docs-mcp-response.txt and docs-llms.txt expose streamable-HTTP read-only documentation search endpoint https://docs.nvidia.com/nvcf/_mcp/server. Not deployment tools.”
SDK packages
Partially supported
“Python workload helpers are official; Go libraries are shared infrastructure; no public multi-language management SDK packages established. docs-overview-container-functions.md and github library inventories.”
Auth documented
Supported
“docs-overview-api.md and CLI describe three bearer types, programmatic generation, scopes per endpoint and expirations.”
Self-serve credentials
Not applicable
“No hosted NVCF signup/API account; operator installs service. Local nvcf-cli init/api-key generate is programmatic and documented, but should not be scored as hosted self-registration.”
Webhooks
Not supported
“No documented signed user-facing event webhook delivery in retained current docs index/API. Internal auth webhooks and GitHub CI are not such a feature.”
CLI tool
Supported
“docs-overview-cli.md describes official install, lifecycle, keys, account/registry administration.”
Error model
Partially supported
“OpenAPI contains documented 403/429 responses and quota runbook shows JSON title/detail-style failure, but no unified public JSON error schema established. Do not preserve old unsupported full score.”
Rate limits documented
Partially supported
“docs-overview-llm-gateway.md documents tokenRateLimit syntax, units and request behavior; spec includes 429. No full Retry-After/rate-limit-header contract established.”
Agent config files
Supported
“github-AGENTS.md plus public user/developer skills in github-ai-tooling-README.md and CLI SKILL. Multiple surfaces, not old one-surface partial.”

05 Coding harnesses Which agents and IDEs are documented to run here

Advertised integrations 0

The public surface names no coding harness as supported or documented-to-run-here.

Plausible compatibility 0

Derived, not stored: the product passes SSH access, so any CLI coding agent is installable inside it. This is plausible compatibility, not a supported integration.

No general shell access documented, so nothing is derived. Absence is not unsupported.

06 Compliance & regions Standards held, with ordinal rigor; never a score

Certifications held 0

No attestations found on the public surface. Empty means nothing is advertised, after checking, not that nothing was checked.

Regions —

No region list captured.

07 Open source Apache-2.0 · fetched by machine, not quoted from marketing

Licence & repository

Licence: Apache-2.0OSI
Primary language: Go
Stars: 237
Forks: 93
Created: 1 Apr 2026
Archived: No

Activity

Last commit: 15 Oct 2026
Latest release: src/control-plane-services/event-ledger/v0.18.0 · 15 Oct 2026
Issues open / closed: 381 / 774 · 67% closed

2919 GitHub releases at fetch; many independently versioned component releases, including several on 2026-10-14 and latest event-ledger v0.18.0 on 2026-10-15. This is not a single umbrella product-version sequence. Default docs release 1.0.3 is docs-only; qualified stacks are Self-Managed 1.0.1, Compute Plane 1.0.0, Observability 1.0.0.

Fetched via gh on 15 Oct 2026.

Self-hosting

Methods: helmkubernetessource-build
You operate: kubernetesnatscassandraopenbaoenvoy-gatewaycert-manager

Apache-2.0 software plus operator infrastructure/operations cost; no hosted pricing tier. CLI local k3d installation or manual Helmfile/Helm stacks; requires GPU Operator, SMB CSI driver and persistent StorageClass. Official artifact inventory uses public nvidia/nvcf NGC images/Helm repository and upstream supporting images. Helmfile expects NVCF charts in OCI: copy public repository charts into the configured OCI registry. Quickstart/image-mirroring retain NGC API-key/org/team instructions; NGC nvcf-onprem credentials are conditional on that distribution route, not a blanket requirement for every public artifact. Issue #12 is still open, but current main contains source, BUILD.bazel and documented build commands for all five named API services and public worker code. Do not claim source absence; complete clean-room build/deployment was not exercised. Optional observability is a separate Early Access stack (default VictoriaMetrics backend); operator owns long-term retention/logging/alerts. Optional UI is unauthenticated and must stay behind trusted network/VPN/authenticating proxy. NvSnap checkpoint/restore is beta and per-cluster opt-in.

Documentation signals 8/8

SupportedQuickstart
SupportedAPI reference
SupportedSelf-host guide
SupportedChangelog
SupportedOpenAPI
Supportedllms.txt
SupportedExamples
SupportedArchitecture doc

08 Company & presence Facts only, not scored

Company

Headquarters: Not disclosed
Founded: NVIDIA founded 1993; NVCF public repository created 2026-04-01
Team size: Not disclosed
Hiring: Unknown

Activity

Last release: src/control-plane-services/event-ledger/v0.18.0 (2026-10-15T00:46:42Z) github.com
Last blog post: — github.com

Presence 4

Links found on the public surface scanned; not an exhaustive search.

09 Sources & method How this entry was produced

Public surface only

Everything here was read from github.com/NVIDIA/nvcf's unauthenticated pages, docs, pricing and repository on 15 Oct 2026. Nothing behind a login was used.

Statuses, not scores

Capabilities are pass, partial or fail against a written rubric (features v21). “Not assessed” means the scan predates the key.

Confidence is separate

Each fact carries its own confidence. Low-confidence facts are shown, never hidden, so you can weigh them.

Schema: scan v5
Features registry: v21
Keys tracked today: 61
Assessed on this entry: 61

Something wrong on this page?

Nearby in the atlas

All Serverless products →