Skip to main content
📖 The AI Tool Bible

AI serverless GPUs

Editorial picks for "serverless gpu".

23 tools

All Fine-tuning →
TA

Together AI

Featured
Fine-tuning · Llama / Mistral / Qwen / DeepSeek and others
8.6

Fine-tune & serve open-weight models (Llama, Mistral, DeepSeek).

Paid· Pay-per-token; fine-tuning per-tokenopen modelsfine-tuning
MO

Modal

Fine-tuning · Infrastructure (any model you can host)
8.7

Serverless GPUs and infra for training & serving ML.

Freemium· $30/mo free credits; pay-as-you-go GPU ratesserverless GPUfine-tuning
RE

Replicate

Fine-tuning · Thousands of community + first-party models
8.5

One-API platform for running and fine-tuning open-source models.

Paid· Pay-per-second of GPU timemodel hostingfine-tuning
CL

ClearML

Agents · Model-agnostic
8.3

End-to-end platform for streamlining AI development and deployment.

Freemium· Community plan is $0; Pro plan is $15 per user/month plus usageExperiment trackingDataset versioning
RU

RunPod

Fine-tuning · Bring-your-own (any open-weight or custom model)
8.3

Paid· Pod H100 SXM: $4.39/hr; Pod RTX 4090: $0.89/hr; Serverless 80 GB H100 PRO: $5.89/hr; Serverless 24 GB 4090 PRO: $1.34/hr; Cluster H100 SXM: $4.79/GPU/hrModel training and fine-tuningServerless inference deployment
VL

vLLM

Fine-tuning · Multi-model (open-weight LLMs: Llama, Qwen, DeepSeek, Mistral, Gemma, Phi, etc.)
8.3

Open-source high-throughput inference engine for serving LLMs with PagedAttention and continuous batching.

Free· Free and open-source (Apache 2.0); self-hosted infrastructure costs applyllm-servingself-hosted-inference
BE

BentoML

Agents · Multi-model
8.2

Open-source framework and managed platform for serving and scaling AI models in production.

Freemium· OSS free (Apache 2.0); managed Bento cloud has free tier + usage-based pricingmodel-servingllm-inference
CO

CoreWeave

Fine-tuning · DeepSeek
8.2

An AI cloud platform purpose-built to turn complex AI workloads into discovery and creation.

Enterprise· Free tier · from $0.015/moAI model trainingAI inference
DA

Daytona

Agents
8.2

Secure and elastic infrastructure for running AI-generated code.

Freemium· Compute vCPU is $0.0504/h; Memory is $0.0162/h per GiB; Storage is $0.000108/h per GiB after the first 5 GiB are free; Windows OS is $0.0858/vCPU/h; $200 in free compute is includedAI evaluationsCode interpreter
SG

SGLang

Fine-tuning · Multi-model (DeepSeek, Qwen, Llama, Mistral, GLM, GPT-OSS)
8.2

Open-source inference framework for large language, vision-language, and diffusion models optimized for agentic workloads and large-scale serving.

FreeAgentic workloadsLarge-scale serving
LA

Lambda

Fine-tuning · NVIDIA VR200 NVL72, NVIDIA GB300 NVL72, NVIDIA HGX B200, NVIDIA HGX B300, NVIDIA H100
8.1

Supercomputers for training and inference

Paid· NVIDIA HGX B200 1-Click Cluster (16 GPUs, 2 weeks – 1 year) is $9.86 per GPU per hour; NVIDIA H100 1-Click Cluster (16 GPUs, 2 weeks – 1 year) is $6.16 per GPU per hour; NVIDIA HGX B200 Instance (8x, 180 GB VRAM) is $7.69 per GPU per hour; NVIDIA HGX H100 Instance (8x, 80 GB VRAM) is $4.59 per GPU per hour; NVIDIA HGX A100 Instance (8x, 80 GB VRAM) is $3.19 per GPU per hour; NVIDIA HGX A100 Instance (8x, 40 GB VRAM) is $1.99 per GPU per hour; NVIDIA Tesla V100 Instance (8x, 16 GB VRAM) is $0.79 per GPU per hourTraining and inferenceTraining foundation models
FA

Fal.ai

Image Generation · Multi-model (Flux, Stable Diffusion, video/audio models)
8.0

A unified API for generative media models with fast inference, training, and production infrastructure.

Paid· Usage-based; serverless from ~$1.89/GPU-hour, per-output pricing on model APIsimage-generationvideo-generation
AN

Anyscale

Fine-tuning · Infrastructure (any model)
7.9

Anyscale helps AI builders run data-intensive workloads to build and deploy Foundation Models and AI at scale on any cloud, powered by Ray.

PaidMultimodal data curationDistributed model training
FA

Fireworks AI

Fine-tuning · Multi-model (DeepSeek, Qwen, GLM, Kimi, Gemma, Minimax, others)
7.9

A training and inference platform that transforms open models into specialized intelligence.

Paid· Serverless: From $0.05 / 1M input tokens; Managed Training: From $0.50 / 1M training tokens; Dedicated Training API: From $8.00 / GPU-hourOpen model inferenceModel training
FE

FedML

Fine-tuning · Bring-your-own (PyTorch, Hugging Face)
7.3

Distributed training, fine-tuning, and serving platform with federated learning roots.

Freemium· Open-source library free; managed GPU usage pay-as-you-gofine-tuningdistributed-training
BE

Beam

Coding · openai/gpt-oss-20b
7.2

Serverless GPU infrastructure for AI workloads with sub-second cold starts and bring-your-own-cloud support.

Freemium· $30 free credit refreshed monthly; usage-based beyond thatgpu-inferenceagent-sandboxes
PG

Paperspace Gradient

Fine-tuning · Bring-your-own (PyTorch, TensorFlow, Hugging Face)
7.2

MLOps platform for building and scaling AI projects on DigitalOcean.

Freemium· Free tier · from $8/moexploring libraries and datasetstraining and fine-tuning models
NR

Neu.ro

Agents · Multi-model
7.1

Infrastructure-agnostic MLOps platform for the full ML/DL lifecycle across public, hybrid, and on-prem clouds.

Enterprise· Contact sales; no public pricingmlopsmodel-training
FO

Forefront

Fine-tuning · Multi-model (Mistral-7B, Mixtral, Phi-2)
7.0

Fine-tune and serve open-source LLMs on your own data without managing GPUs.

Paidfine-tuningopen-source-llms
VE

Velda

Fine-tuning
6.7

Serverless GPU Cloud without overhead of containers or cluster management.

Freemium· Free plan; shell pool: $0.40 per hour for 4 vCPU, 16 GB RAMTraining and inference jobsServing and data pipelines
E2

E2B

Agents · LLM-agnostic (works with GPT-4o, Claude, Mistral, Llama, and self-hosted models)

E2B runs the isolated machine for AI agents to execute code and use tools.

Freemium· Free tier · from $150/moCoding and computer use agentsReinforcement learning rollouts
SC

Scalattice

Agents · Open-weight models including Qwen 2.5 Coder, Llama 3.3 70B, DeepSeek R1, Gemma

Pay less for AI and earn from your GPU.

Freemium· $0 platform fee to connect a machineManaged production inference for developersEarning token revenue on inference-only catalog jobs
X4

X402vps

Agents

Pay-per-hour Docker containers for autonomous AI agents, billed in USDC via the x402 protocol

Paid· Usage-based hourly: ATOM $0.01/hr (0.1 vCPU, 256MB RAM, 2GB disk, no internet); CELL $0.02/hr (0.25 vCPU, 512MB, 5GB, 100MB/hr bandwidth); NODE $0.04/hr (0.5 vCPU, 1GB, 10GB, 500MB/hr); CORE $0.08/hr (1 vCPU, 2GB, 20GB, 2GB/hr). Bandwidth overage $0.01/GB. Exec calls are free. Paid in USDC on Base.Ephemeral web scraping sandboxesLLM code-interpreter backends