Modal
InfrastructureModal (Modal Labs) is a New York-based, Python-native serverless cloud for AI workloads. It offers on-demand GPU/CPU compute, multi-node clusters with RDMA, managed LLM inference endpoints, and fast-booting sandboxes, including full VM sandboxes, which Modal says scale to 1M concurrent. It sells execution infrastructure, not RL environments. AI labs, agent companies and RL-data vendors use it to run RL rollouts, post-training and large fleets of parallel sandboxed environments. It integrates with Anthropic's Claude Managed Agents and Claude Science, Cognition's Devin Outposts and the OpenAI Agents SDK, and ships an open-source RL/SFT library (Modal Dojo).
What's new
- 1 Oct 2026product
Runtime by Modal: VM Sandboxes, Modal Clusters and Sticky Sessions reach GA; Endpoint Candidates enters private beta
Announced at Modal's Runtime conference in San Francisco. VM Sandboxes give agents a full Linux VM, with Docker-in-Docker, databases and graphical environments; Linear, Legora and Snorkel are named as users. Modal Clusters (GA) provide multi-node gang scheduling with RDMA up to 6.4 Tbps; Modal says Decagon runs post-training, 1X pre-training and Runway inference on them. Sticky Sessions (GA) send a client to the same container across requests. Endpoint Candidates (private beta) mirror production traffic to test alternative serving setups. The roundup lists Sandbox Sidecars as GA, but Modal's dedicated Sidecars post from the same day says beta.
Why it matters: Docker-capable VM sandboxes and multi-node RL training are now generally available. Snorkel, an RL-environment and data vendor, is named as running millions of agent simulations a month on Modal. All customer names are vendor claims.
- 1 Oct 2026product
Sandbox Sidecars: a trust boundary next to agent sandboxes
Sidecars are isolated containers on the same host as a Sandbox. They can inject credentials, proxy traffic, redact data and monitor the agent. Modal reports a p50 latency of 0.58 ms per call between a Sandbox and its Sidecar, against 2.6 ms between separate Sandboxes. Ramp is named as running its own egress proxy for its coding agent. This post calls Sidecars beta.
Why it matters: Keeps credentials and harness logic away from code the agent generates, which matters for RL and agent environments that call outside services.
- 28 Sep 2026funding
TechCrunch: Modal close to raising $750M at a $15.75B valuation, led by Accel
A single-source report that the round is nearly closed. The valuation would be more than 3x the $4.65B Series C from May 2026. TechCrunch also gives about 150 employees and names Cognition, Suno, Ramp and Substack as customers.
Why it matters: A strong sign of scale and staying power for a vendor that runs RL rollouts and sandboxes. Modal has not confirmed it.
- 24 Sep 2026open-source
Quail: open-source AI-SQL inference engine built with CMU's Full Stack Data Lab
A collaboration between 'inference researchers at Modal' and CMU database researchers, written up by Charles Frye and Shreya Shankar. Modal reports more than 1B tokens per minute per H100 and 1.84x the speed of vLLM (geometric mean across tasks on its AI-SQL benchmark). Code is at github.com/fsdatalab/quail.
Why it matters: Shows Modal publishing inference research with academic partners, beyond its infrastructure work.
- 23 Sep 2026research
Engineering write-up: serving trillion-parameter Kimi K2.6 at trillions of tokens
Covers DFlash speculative decoding, TP4 tensor parallelism, FP8 KV cache and NVFP4 quantization, HiCache, and load-aware routing. Modal reports 2.8x per-user speed and 5.6x throughput. It says it serves hundreds of billions of tokens a day, including through OpenRouter and Vercel AI Gateway.
Why it matters: Modal is now a large-scale LLM inference provider, not only a place to run compute.
- 15 Sep 2026product
New controls for unauthenticated web endpoints
Endpoints can be filtered and tagged by authentication status. Deploying an unauthenticated URL now writes an audit-log event, and a workspace setting can block web endpoints from being deployed without proxy auth. RBAC user groups (with SCIM sync) and SSO metadata-file upload shipped the same day.
Why it matters: This hardening follows the July incident, which involved a customer's unauthenticated endpoint. Admins can now enforce authentication across a workspace.
- 14 Sep 2026product
Product update: Sandbox Sidecars (public alpha), VM Sandboxes (public beta), filesystem API and directory snapshots GA, new open-weight model endpoints
Day-zero endpoints for Kimi K3, Qwen 3.8-2.4T-A95B, GLM-5.3 and GLM-5.3-Flash. Also: environment budgets, a regional pricing multiplier cut from 1.5x to 1.15x, Private Environments and Regional Proxy selection. Python SDK 1.5.5 and JS/Go SDK 0.10.0. Announced Runtime for 2026-10-01 in San Francisco. VM Sandboxes went GA on 2026-10-01.
Why it matters: Multi-container sandboxes and Docker-in-sandbox let teams run more complex RL and agent environments, such as an app plus its database.
- 2 Sep 2026hire
Modal opens a London office and hires a VP EMEA
London is Modal's second European office after Stockholm, which has an engineering team of nearly 20. Hugh Killingbeck-Jones joins as VP EMEA. Modal names Black Forest Labs and Legora as European customers and says it plans more data-residency and compliance support.
Why it matters: EU and UK buyers get local sales and support. Data residency is still on the roadmap.
- 14 Aug 2026product
Changelog: Sandbox filesystem API GA; V2 Sandboxes opt-in (directory snapshots GA on 2026-08-12)
The filesystem API is GA, and Modal is speeding up deprecation of the legacy file API. Directory snapshots went GA on 2026-08-12, and V2 Sandboxes became opt-in (MODAL_SANDBOX_V2=1) on 2026-08-14, both per modal.com/changelog. The public changelog's earliest entry is dated 2026-08-06 (environment budgets). Regional proxies followed on 2026-08-31.
Why it matters: GA file I/O and snapshots matter for saving and resuming environment state in RL rollouts.
- 29 Jul 2026legal
Modal statement on the Hugging Face agent incident
Hugging Face disclosed an agent intrusion in which an attacker abused a Modal customer's public endpoint that ran user-submitted code in Modal Sandboxes. Modal says its platform and isolation were not compromised and no other customers were affected. It recommends authentication, IP allowlists, egress restrictions and treating all input as untrusted.
Why it matters: A reference point for anyone exposing sandboxes publicly. Modal attributes the incident to the customer's configuration, not to a failure of its isolation.
- 21 Jul 2026partnership
Modal is a launch partner for Cognition's Devin Outposts and open-sources modal-devin
Devin sessions can run inside a customer's own Modal Sandboxes, including GPU-backed ones. Set up with `uvx modal-devin init`. modal-devin is open source.
Why it matters: Deepens the Cognition relationship. Devin's actions run in sandboxes the customer controls.
- 16 Jul 2026product
Rebuilt sandbox scheduler runs 1 million concurrent sandboxes (beta)
The new scheduler has no central database in the scheduling path. Modal reports creating 1M sandboxes in under 60 seconds, tens of thousands per second, with median startup under 500 ms. The new system is in beta and will eventually handle all sandbox scheduling.
Why it matters: This is the vendor's own scale claim, and it bears directly on large parallel RL rollouts.
- 15 Jul 2026partnership
Day-zero hosting of Thinking Machines' Inkling model as a Managed Endpoint
Modal partnered with Thinking Machines on day-zero support, with token-based pricing and custom DFlash speculative decoding. Modal reports 250 tokens per second per user, 67% faster than the model's built-in speculative path.
Why it matters: A launch collaboration with a frontier lab.
- 1 Jul 2026research
Multi-token Residual Prediction (MRP) paper co-authored by Modal Research
Speeds up diffusion language models (arXiv 2605.18817). The Modal blog lists Yufeng Xu and Rahul Chalamala as Modal Research, with Yucheng Lu of NYU Shanghai's HeavyBall Research. The blog reports up to 1.56x throughput in SGLang; the arXiv abstract reports up to 1.4x lossless speedup, with +22.6 points on HumanEval and +17.7 on GSM8K under aggressive decoding.
Why it matters: Modal has a named research group that co-authors peer-style papers.
Founding team
What practitioners should know
For researchers
- Modal has a named research group: the MRP paper's co-authors are listed as 'Modal Research'. Its published inference work includes Quail (with CMU), the FlashAttention-4 PRs, DFlash speculators (with Z Lab and SGLang), the trillion-parameter serving write-up, and the LLM Engineer's Almanac benchmarks with open code (modal-labs/stopwatch).
- RL tooling: Modal Dojo (formerly Training Gym, MIT) defines RL or SFT jobs from a reward function, model and environment. Modal says it supports customers doing SFT and RL on GLM 4.7 and Kimi K2.6. AE Studio published a reproducible Lean theorem-proving RL case study (ES vs GRPO).
- Open releases: modal-dojo, modal-auto-research-skills, modal-devin, Quail and the DFlash speculators. The core platform is proprietary.
- RL use is published as vendor case studies: Applied Compute (rollouts, evals and inference with thousands of parallel sandboxes), Snorkel (millions of agent simulations a month on VM Sandboxes) and Decagon (post-training on Modal Clusters).
For program managers
- Modal sells infrastructure for running RL environments. It does not sell environments or data.
- Frontier-lab ties: two Anthropic integrations (Claude Managed Agents, May 2026; Claude Science, June 2026), day-zero hosting of Thinking Machines' Inkling (July 2026), and launch partner for Cognition's Devin Outposts (July 2026). The baseline's Meta/CWM sandbox usage is still self-claimed.
- Scale signals: $355M Series C at $4.65B (May 2026). A ~$750M round at $15.75B led by Accel was reported by TechCrunch on 2026-09-28 from a single source and is unconfirmed. Annualized revenue was over $300M in May 2026, and TechCrunch puts headcount at about 150.
- Offices in NYC (HQ), SF, Stockholm and London; London opened in September 2026 with a VP EMEA. Modal acquired Butter in April 2026 to strengthen its sandbox team. Modal's first Runtime conference was held on 2026-10-01.
For data ops
- SOC 2 Type 2, with the report on trust.modal.com. HIPAA via BAA on the Enterprise plan. The security docs list no ISO 27001 or GDPR certification.
- Isolation uses gVisor, with an optional VM runtime; VM Sandboxes went GA on 2026-10-01. Function inputs and outputs are kept for at most 7 days.
- July 2026 Hugging Face incident: a customer's public, unauthenticated code-execution endpoint was abused. Modal says its isolation held. In September it added controls to block unauthenticated endpoints workspace-wide, with audit-log events and tagging.
- Enterprise controls added August-September 2026: environment budgets, RBAC user groups with SCIM sync, SSO IdP XML upload, regional proxies, Private Environments, and a public changelog at modal.com/changelog. Sidecars can sit in front of agent egress as a proxy.
- Integrations: Python, JS/TS and Go SDKs; Claude Managed Agents; Claude Science; Devin Outposts; the OpenAI Agents SDK sandbox extension (April 2026).
Benchmarks & research
Products
Scale & velocity
Research depth
Capital
| Date | Round | Amount | Led by | |
|---|---|---|---|---|
| 28 Sep 2026 | undisclosed unsettled | $750M valuation $15.75B | Accel | source ↗ |
| Unannounced new round (stage not disclosed) · $750M (reported as nearly closed; Modal has not announced it) | ||||
Security & compliance
Product
Buyer analysis
How we verified this
Related vendors
Sources
- techcrunch.com/2026/09/28/source-inference-provider-modal-labs-closing-i · 2026-10-05, Reported ~$750M round at $15.75B led by Accel (not closed); ~150 employees; customers Cognition, Suno, Ramp, Substack
- modal.com/company · 2026-10-05, Offices (NYC, SF, Stockholm, London); 30+ roles; investors GC, Redpoint, Lux, Amplify, Creandum; $466M+ raised; founders
- modal.com/sitemap.xml · 2026-10-05, Used to list every blog and changelog post in the window
- modal.com/blog/runtime-product-update-sandbox-endpoints · 2026-10-05, Runtime 2026-10-01: VM Sandboxes, Sidecars, Clusters and Sticky Sessions GA; Endpoint Candidates private beta
- modal.com/blog/vm-sandboxes-agent-computers · 2026-10-05, VM Sandboxes GA; Linear, Legora, Snorkel
- modal.com/blog/modal-clusters-generally-available · 2026-10-05, Modal Clusters GA; RDMA 6.4 Tbps; Decagon, 1X, Runway
- modal.com/blog/introducing-sandbox-sidecars-trust-boundary · 2026-10-05, Sidecars (labelled beta in this post); Ramp egress proxy
- modal.com/blog/modal-europe-expansion · 2026-10-05, London office, VP EMEA Hugh Killingbeck-Jones, European customers
- modal.com/blog/a-note-on-the-hugging-face-agent-incident · 2026-10-05, Modal's statement on the security incident
- modal.com/changelog/new-controls-for-unauthenticated-web-endpoints · 2026-10-05, Unauthenticated endpoint controls, 2026-09-15
- modal.com/changelog · 2026-10-05, Changelog index: environment budgets (08-06), directory snapshots GA (08-12), V2 opt-in (08-14), regional proxies (08-31), RBAC groups and SSO XML (09-15)
- modal.com/changelog/sandbox-filesystem-api-is-now-generally-available · 2026-10-05, Sandbox filesystem API GA, 2026-08-14
- modal.com/blog/scaling-to-1-million-concurrent-sandboxes-in-seconds · 2026-10-05, 1M concurrent sandboxes on the rebuilt scheduler (beta)
- modal.com/blog/devin-outposts-run-devin-in-modal-sandoxes · 2026-10-05, Devin Outposts launch partner; modal-devin
- modal.com/blog/introducing-claude-managed-agents-with-modal-sandboxes · 2026-10-05, Anthropic Claude Managed Agents integration; Mason AI, DoorDash, Blend
- modal.com/blog/modal-integration-brings-scalable-compute-to-claude-scien · 2026-10-05, Anthropic Claude Science integration, 2026-06-30
- modal.com/blog/inkling-by-thinking-machines-labs-now-available-on-modal · 2026-10-05, Thinking Machines Inkling day-zero support
- modal.com/blog/butter-is-joining-modal · 2026-10-05, Butter acquisition
- modal.com/blog/applied-compute-reinforcement-learning · 2026-10-05, Applied Compute RL case study
- modal.com/blog/reinforcement-learning-infrastructure-problem · 2026-10-05, RL infrastructure essay; Modal Training Gym announcement
- github.com/modal-projects/training-gym · 2026-10-05, Redirects to modal-projects/modal-dojo (MIT; RL and SFT)
- modal.com/blog/building-an-rl-theorem-proving-workflow-on-modal · 2026-10-05, AE Studio Lean RL case study (ES vs GRPO)
- modal.com/blog/building-with-modal-and-the-openai-agent-sdk · 2026-10-05, OpenAI Agents SDK sandbox extension (no formal partnership stated)
- modal.com/blog/physical-intelligence-runs-real-time-remote-inference-for · 2026-10-05, Physical Intelligence case study
- modal.com/blog/product-updates-sidecars-models-dashboard · 2026-10-05, September 2026 product updates
- modal.com/blog/product-updates-rtx-pro-6000-command-k-sandbox-fs-api-and · 2026-10-05, April 2026 product updates; Runway, Imbue, Doppel
- modal.com/blog/introducing-auto-endpoints · 2026-10-05, Auto Endpoints launch
- modal.com/blog/achieve-sota-specdec · 2026-10-05, Speculative decoding write-up; DFlash speculators with Z Lab and SGLang
- modal.com/blog/flash-attention-4-faster · 2026-10-05, FlashAttention-4 contributions
- modal.com/blog/trillion-tokens-trillion-parameters · 2026-10-05, Kimi K2.6 inference engineering write-up
- modal.com/blog/quail-billion-tpm · 2026-10-05, Quail open-source engine with CMU
- modal.com/blog/multi-token-residual-prediction · 2026-10-05, MRP post; authors listed as Modal Research
- arxiv.org/abs/2605.18817 · 2026-10-05, MRP paper (submitted 2026-05-12)
- modal.com/blog/autoscaling-autoresearch · 2026-10-05, Autoresearch post and open-source skills
- modal.com/blog/modal-series-c · 2026-10-05, Series C details, investors, customers
- modal.com/llm-almanac · 2026-10-05, LLM Engineer's Almanac benchmarks
- modal.com/gpu-glossary · 2026-10-05, GPU Glossary
- modal.com/docs/guide/security · 2026-10-05, SOC 2 Type 2, HIPAA BAA, gVisor and VM runtime, 7-day I/O retention, trust portal
- github.com/erikbern · 2026-10-05, Erik Bernhardsson's verified GitHub (bio cites Modal and Spotify) and headshot avatar
- erikbern.com/about.html · 2026-10-05, Erik Bernhardsson bio
- modal.com/ · 2026-06-07, Official homepage, serverless AI infrastructure positioning, GPU autoscaling, inference, fine-tuning
- modal.com/blog/announcing-our-series-b · 2026-06-07, Official Series B, $87M, $1.1B valuation, Lux Capital lead, $111M total, founders, customers
- modal.com/blog/soc2type2 · 2026-06-07, Official, SOC 2 Type II compliance
- modal.com/blog/hipaa · 2026-06-07, Official, HIPAA-compliant workloads via BAA on Enterprise plan
- techcrunch.com/2026/02/11/ai-inference-startup-modal-labs-in-talks-to-ra · 2026-06-07, TechCrunch, founded 2021, ~$50M ARR (Feb 2026), $2.5B talks, CEO background
- www.datacenterdynamics.com/en/news/modal-labs-secures-funding/ · 2026-06-07, DCD, $355M Series C, $4.65B, two tranches ($2.5B/$4.65B), ~$300M ARR up from ~$60M
- siliconangle.com/2026/05/21/serverless-ai-infrastructure-startup-modal-l · 2026-06-07, SiliconANGLE, Series C close date ~2026-05-21
- techstartups.com/2026/05/21/modal-labs-raises-355m-quadrupling-valuation · 2026-06-07, Tech Startups, Series C details, investors, customers
- github.com/modal-labs · 2026-06-07, GitHub org, client SDKs only (Python Apache-2.0, JS/TS/Go libmodal MIT, examples MIT)
- github.com/modal-labs/modal-client · 2026-06-07, Modal client SDK repo
- www.linkedin.com/company/modal-labs · 2026-06-07, LinkedIn public snippet, ~153 employees (April 2026), NYC, founded ~2021