Skip to content

Baseten

Dedicated inference for open-weight models.

Baseten runs open-weight models on dedicated infrastructure with an OpenAI-compatible endpoint. Its catalog is deliberately small — a curated set of frontier open models served fast, rather than an aggregation of everyone else's.

Open-weight focus Dedicated deployments Small curated catalog
Status
Live
Models in catalog
16
Text / multimodal
Publishes pricing for
No prices published
Largest context window
Model id style
Canonical `provider/model` for open-weight models.
Billing
Usage-based, billed by Baseten.
Catalog last checked
Website
baseten.co
Documentation
Developer docs
Availability data only in this snapshot. This build was generated from FreeRouter's model-map snapshot, which records which gateway serves which model but carries no pricing, context windows, or descriptions. Those columns fill in once the catalog refresh runs against the gateway APIs — see about the data.
Catalog

16 models on Baseten

The 16 most widely available are shown; the rest are searchable in the full model list. "Id on this gateway" is what Baseten calls the model — where it differs from the canonical id, that difference is what a router has to translate.

Search all 16 models on Baseten
ModelVendorTypeId on this gatewayGatewaysContext
gpt-oss-120b openai/gpt-oss-120b OpenAI Unclassified same as canonical 5
inkling thinkingmachines/inkling Thinkingmachines Unclassified same as canonical 4
kimi-k2.6 moonshotai/kimi-k2.6 Moonshot AI Unclassified same as canonical 3
kimi-k2.7-code moonshotai/kimi-k2.7-code Moonshot AI Unclassified same as canonical 3
kimi-k3 moonshotai/kimi-k3 Moonshot AI Unclassified same as canonical 3
inkling-small thinkingmachines/inkling-small Thinkingmachines Unclassified same as canonical 3
deepseek-v4-flash-0731 deepseek-ai/deepseek-v4-flash-0731 Deepseek Ai Unclassified same as canonical 1
deepseek-v4-pro deepseek-ai/deepseek-v4-pro Deepseek Ai Unclassified same as canonical 1
deepseek-v4-pro-0813 deepseek-ai/deepseek-v4-pro-0813 Deepseek Ai Unclassified same as canonical 1
nvidia-nemotron-3-ultra-550b-a55b nvidia/nvidia-nemotron-3-ultra-550b-a55b NVIDIA Unclassified same as canonical 1
glm-4.7 zai-org/glm-4.7 Z.ai Unclassified same as canonical 1
glm-5.2 zai-org/glm-5.2 Z.ai Unclassified same as canonical 1
glm-5.2-fast zai-org/glm-5.2-fast Z.ai Unclassified same as canonical 1
glm-5.3 zai-org/glm-5.3 Z.ai Unclassified same as canonical 1
glm-5.3-fast zai-org/glm-5.3-fast Z.ai Unclassified same as canonical 1
glm-5.3-flash zai-org/glm-5.3-flash Z.ai Unclassified same as canonical 1
By model type

What Baseten carries

A gateway with 2,000 models is not necessarily a gateway with 2,000 chat models.

Not exclusive

6 of these models are on other gateways too

Anything in this list can be served from somewhere other than Baseten without changing the model id your application sends.

ModelVendorGatewaysContext
gpt-oss-120b openai/gpt-oss-120b OpenAI 5
inkling thinkingmachines/inkling Thinkingmachines 4
kimi-k2.6 moonshotai/kimi-k2.6 Moonshot AI 3
kimi-k2.7-code moonshotai/kimi-k2.7-code Moonshot AI 3
kimi-k3 moonshotai/kimi-k3 Moonshot AI 3
inkling-small thinkingmachines/inkling-small Thinkingmachines 3