Modal
NativeAI NativePrivateUpdated November 15, 2026The serverless GPU platform that ML engineers actually want to use.
Try it free →- Website
- modal.com
- Founded
- 2021
- CEO
- Erik Bernhardsson
- HQ
- New York, NY, United States
- Stage
- Series C
- Revenue
- Not disclosed
- Total Funding
- $466M
- Employees
- 501-1000
Executive Summary
Updated Nov 15, 2026
What they do
- Serverless GPU and CPU compute platform driven entirely from Python code
- Offers inference, training, batch jobs and secure sandboxes for AI agents
- Pools NVIDIA B200, H200, H100 and A100 capacity across multiple clouds
Why it matters
- Says annualized run rate passed $1B in October 2026, with 15x yearly growth
- Sandboxes bring in nearly a third of revenue as AI agent workloads grow
- Customers include Lovable, Ramp, Cognition, DoorDash, Suno and Substack
Best fit
- ML teams serving bursty, low-latency inference without managing GPU clusters
- Agent and coding-tool builders running untrusted AI-generated code at scale
- Companies fine-tuning models or running large parallel batch and RL jobs
Watch-outs
- Python-native SDK and runtime create real platform lock-in
- Per-second consumption billing can make sustained heavy workloads costly
- Very fast growth may strain support, capacity and enterprise readiness
Products & Offerings
Modal Platform
Serverless container runtime that runs Python functions on GPUs or CPUs, with sub-second cold starts and scale-to-zero.
Modal Sandboxes
Isolated, gVisor-based ephemeral environments for running AI-generated code, coding agents and RL rollouts at very high concurrency.
Modal Inference
Autoscaling serving for LLM and multimodal models, with vLLM support and OpenAI-compatible endpoints.
Modal Training
Fine-tuning and multi-node training up to 128 B200 GPUs on InfiniBand, plus parallel hyperparameter sweeps.
Modal Batch
Massively parallel batch processing of data and model jobs on per-second-billed serverless compute.
Latest News
LMSYS Org · 3 days ago
Positivecitybiz · 6 days ago
NeutralFourWeekMBA · 7 days ago
PositiveHoodline · 8 days ago
NeutralBigGo Finance · 8 days ago
NeutralUpdated Nov 15, 2026
Modal is now a leading serverless AI compute platform. It reports more than $1B in annualized run rate, well-known AI-native customers and $1.22B in total funding. Teams building inference, fine-tuning or agent-sandbox workloads should shortlist it against Baseten, RunPod, Together AI and the GPU services of the big clouds. Before committing sustained heavy workloads, benchmark cost at steady-state usage and confirm compliance, data residency and capacity guarantees.
Signal Provenance
Latest observation date and source for each signal category
| Signal Category | Last Updated | Source |
|---|---|---|
Funding Signals Last funding round, Funding stage | Nov 12, 2026 | Cited web research |
Hiring Velocity Headcount, Headcount growth | Nov 15, 2026 | Research agents (estimate) |
Technical Output Research papers, GitHub activity | Nov 13, 2026 | Hugging Face Hub + news feed + research agents |
News Sentiment News mentions, News sentiment | Nov 15, 2026 | Verified news feed |
Partnership Signals Enterprise partnerships, Distribution partners | Nov 12, 2026 | Research agents (estimate) |
Revenue Indicators ARR estimate, Revenue growth | Nov 15, 2026 | Research agents (estimate) |
Leadership Signals CEO tenure, Founder-led | Nov 15, 2026 | Research agents (estimate) |
Competitive Position Category rank, Market share | Nov 7, 2026 | Research agents (estimate) |
Why this Bluescore?
6 dimensions, each scored 0–10 against all 250 companies. Each adds its weight × score to the Bluescore — see methodology
Modal's Bluescore of 47.4 rests most on Talent & Research (7.5, top 20%) and Customer Adoption (5.6, 71st percentile). Its weakest area is Business Model & Revenue (3.0, 9th percentile).
In plain English
Modal's Bluescore of 47.4 out of 100 is lifted mainly by Talent & Research, where its score of 7.5 of 10 puts it at the 80th percentile among the AI250 and adds 12.0 of 16 points, with Customer Adoption also helping at 5.6 of 10 (71st percentile). The biggest drag is Business Model & Revenue, which carries 21% weight but scores only 3.0 of 10 (9th percentile) and adds just 6.2 of 21 points, leaving 14.8 points unclaimed, while Data & Model Moat is also weak at 3.5 of 10 (22nd percentile). Improving Business Model & Revenue would move the score more than anything else, because it combines a large weight with the most points still available to gain.
Written by AI from the scores below and checked against them; updated nightly.
Traction Signals
Last Round
Series C
Risks
Updated Nov 15, 2026