
Baseten Inference Stack
Baseten Inference Stack for high performance with custom kernels, decoding, caching on optimized infra.
- Currently Offered

Baseten Inference Stack for high performance with custom kernels, decoding, caching on optimized infra.
The Baseten Inference Stack combines runtime, kernel and routing optimizations for apps.
It delivers low latency high throughput with bleeding edge performance research, custom kernels, latest decoding techniques and advanced caching. Supports multi cloud and self hosted. Source: baseten.co platform and model performance pages.


Fireworks AI is a managed inference and fine-tuning platform for open-source generative AI models.



Hedra is a model development service that pairs customers with forward-deployed researchers, research agents, and GPU infrastructure.


Together AI is a serverless and dedicated inference API for open-source AI models with OpenAI-compatible endpoints.

Fireworks AI is a managed inference and fine-tuning platform for open-source generative AI models.

Inference is a model and serving stack tuning service that tunes both together for each customer's workload and GPUs.

Hedra is a model development service that pairs customers with forward-deployed researchers, research agents, and GPU infrastructure.

Together AI is a serverless and dedicated inference API for open-source AI models with OpenAI-compatible endpoints.

World Model Runtime is an inference engine for video and world models.



SkyPilot Tokens is an inference API serving open models across verified GPU fleets for AI developers.


Model Deployments manages immutable model releases with promotion, rollback, logs and analytics.


Cloudflare AI: serverless inference, agents, RAG, and secure runtime primitives on the global edge.


Serverless Inference serves open models through an OpenAI-compatible API billed by token use.


IonRouter is a high-throughput, low-cost inference API and drop-in OpenRouter alternative powered by Cumulus Labs' IonAttention engine.


Unified production inference platform with gateway, router, cache, observability, evaluation, fine-tuning, and custom GPU hosting.


AI inference platform delivering serverless and dedicated GPU inference with top-ranked speed and price-performance.


Hosted gateway for routing and governing traffic across hosted and self-hosted AI models.


An integrated compute, training, inference, and sandbox stack for enterprises to train and deploy their own AI models.


Dedicated Endpoints serves one model on GPU capacity reserved for an organization.


Open source AI gateway routing every model behind one zero-markup OpenAI-compatible endpoint.


Inference Neocloud delivers dedicated ASIC racks for high-speed AI inference workloads.


Hyperbolic Inference serves open-source text and vision models to customers who send OpenAI-compatible chat completion requests and pay per token.



InferenceX continuously benchmarks LLM inference throughput, latency, and cost across GPU vendors.



Weave Router is a drop-in proxy that routes each inference request to the lowest-cost model that answers at frontier parity.


Crusoe Managed Inference deploys and scales production LLM workloads with breakthrough speed and throughput.


Evaluations compares models on standard benchmarks or custom tasks with repeatable runs.
SkyPilot Tokens is an inference API serving open models across verified GPU fleets for AI developers.

Model Deployments manages immutable model releases with promotion, rollback, logs and analytics.

Cloudflare AI: serverless inference, agents, RAG, and secure runtime primitives on the global edge.

Serverless Inference serves open models through an OpenAI-compatible API billed by token use.

IonRouter is a high-throughput, low-cost inference API and drop-in OpenRouter alternative powered by Cumulus Labs' IonAttention engine.

Unified production inference platform with gateway, router, cache, observability, evaluation, fine-tuning, and custom GPU hosting.

AI inference platform delivering serverless and dedicated GPU inference with top-ranked speed and price-performance.

Hosted gateway for routing and governing traffic across hosted and self-hosted AI models.

An integrated compute, training, inference, and sandbox stack for enterprises to train and deploy their own AI models.

Dedicated Endpoints serves one model on GPU capacity reserved for an organization.

Open source AI gateway routing every model behind one zero-markup OpenAI-compatible endpoint.

Inference Neocloud delivers dedicated ASIC racks for high-speed AI inference workloads.

Hyperbolic Inference serves open-source text and vision models to customers who send OpenAI-compatible chat completion requests and pay per token.

An open-source AI gateway and proxy that calls many large language model APIs in the OpenAI format.

InferenceX continuously benchmarks LLM inference throughput, latency, and cost across GPU vendors.

Managed GPU cloud for inference, fine-tuning, and prompt workflows.

Weave Router is a drop-in proxy that routes each inference request to the lowest-cost model that answers at frontier parity.

Crusoe Managed Inference deploys and scales production LLM workloads with breakthrough speed and throughput.

Evaluations compares models on standard benchmarks or custom tasks with repeatable runs.
