Modal review

Modal is a different kind of inference platform: instead of mainly calling a hosted model, you deploy your own Python code and Modal runs it on GPUs that start on demand and bill per second. Rates run from $0.000164 per second for an Nvidia T4 to $0.001097 for an H100 SXM5 and $0.001972 for a B300, and you never pay for idle resources. The $0 Starter plan includes $30 a month in compute credits, 3 seats and 10 GPU concurrency, and the Team plan costs $250 a month plus compute with $100 in credits and 50 GPU concurrency. Modal also sells per-token model endpoints for models such as Kimi K3 and GLM 5.3, plus sandboxes and training. It scores lower on ease of use than the serverless APIs above because you own the serving code, but for custom models, multi-step pipelines or unusual architectures that control is the point.

Who it is for

Engineers deploying custom models and pipelines in Python

Price

Starter plan $0 plus compute with $30 per month in free credits; GPUs billed per second from $0.000164 for an Nvidia T4; Team plan $250 per month plus compute.

Nvidia T4 $ 0.000164
Quoted from the vendor pricing page on November 12, 2026.

Key features

• GPUs from Nvidia T4 to B300 billed per second, H100 SXM5 at $0.001097 per second
• No charge for idle resources, only actual compute time
• Model API endpoints for Kimi K3, GLM 5.3, DeepSeek V4.1 Flash and Muse Spark Open billed per token
• Sandboxes for coding agents and untrusted code, plus training on the same GPUs
• Starter plan includes $30 per month in credits, 3 seats and 10 GPU concurrency

Strengths and weaknesses

• Strength: $30 per month in free credits on the $0 Starter plan
• Strength: Per-second billing with scale to zero for spiky inference
• Strength: Full control over the serving stack for custom or fine-tuned models
• Weakness: You write and maintain the serving code for custom models
• Weakness: Starter plan caps GPU concurrency at 10; the Team plan costs $250 per month plus compute
• Weakness: Its per-token endpoint list is shorter than Together AI's or DeepInfra's

Key facts

• H100 SXM5: $0.001097 per second
• Free credits: $30 per month on Starter
• Team plan: $250 per month plus compute, $100 in monthly credits, 50 GPU concurrency
• Billing: Per second of CPU, GPU and memory use

Where it ranks

Modal is No. 8 in our Best AI Inference Platforms verdict.