Who builds the models
60 labs, and the price range across everything each one ships. A provider trains the weights; a gateway runs them for you. Confusing the two is how people end up overpaying.
OpenAI
61 models · US
Creator of the GPT series and the de facto shape of the chat-completions API that most other providers now imitate.
Qwen
41 models · CN
Alibaba's model family, one of the most widely fine-tuned open-weight lineages on the Hub.
30 models · US
Gemini models, served through both the consumer-facing AI Studio API and Vertex AI on Google Cloud.
Mistral AI
20 models · FR
European lab shipping both open-weight and commercial models, with EU-resident inference.
Z.ai
16 models · CN
The GLM series, open-weight and aggressively priced for coding workloads.
Anthropic
15 models · US
Builder of the Claude family, focused on long-context reasoning, tool use and coding agents.
DeepSeek
14 models · CN
Open-weight reasoning and coding models priced far below Western equivalents, with off-peak discounts.
MiniMax
9 models · CN
Long-context text, speech and video models with an OpenAI-compatible API.
Meta
8 models · US
Publisher of the Llama open-weight models, which most inference gateways host directly.
Meta Llama
8 models · US
The Llama open-weight release line - downloadable weights served by dozens of competing hosts.
Moonshot AI
8 models · CN
The Kimi models, strong on agentic and long-document work.
NVIDIA
8 models · US
Nemotron open models, tuned for throughput on NVIDIA inference stacks.
Tencent
7 models · CN
The Hunyuan models, including open-weight releases.
xAI
7 models · US
The Grok family, with large context windows and native search grounding.
AionLabs
6 models
ByteDance Seed
6 models · CN
ByteDance's Seed research line of open-weight models.
Cohere
6 models · CA
Enterprise-focused Command, Embed and Rerank models built around RAG.
OpenRouter
6 models · US
OpenRouter's own routing models, which pick a downstream model per request.
Amazon
5 models · US
The Nova family, served through Amazon Bedrock alongside third-party models.
Perplexity
5 models · US
Sonar models with live web grounding built into the completion itself.
Xiaomi
5 models
Sakana
4 models
Inception
3 models · US
Diffusion-based language models built for very high output throughput.
inclusionAI
3 models
Nous Research
3 models · US
Open, community-driven fine-tunes of leading open-weight bases.
StepFun
3 models
TheDrummer
3 models
Upstage
3 models · KR
The Solar models, optimised for a strong quality-to-size ratio.
Inference.net
2 models
Liquid AI
2 models · US
Liquid Foundation Models, designed for on-device and edge deployment.
Moondream
2 models
Morph
2 models
Nex AGI
2 models
Perceptron
2 models
Poolside
2 models
PrismML
2 models
Reka AI
2 models · US
Natively multimodal models handling video and audio input.
Relace
2 models
Sao10K
2 models
ScaleDown
2 models
Thinking Machines
2 models
Pareto 26.10
2 models
Magnum v4 72B
1 model
ByteDance
1 model · CN
The Doubao and Seed model lines.
Venice
1 model
Dots Studio
1 model
Elorian
1 model
Fireworks
1 model
MythoMax 13B
1 model
IBM Granite
1 model · US
Apache-licensed Granite models aimed at regulated, on-premise deployments.
Mancer
1 model
Meituan
1 model
Microsoft
1 model · US
The small, permissively licensed Phi models, plus Azure-hosted frontier models.
Particle AI
1 model
Recursive
1 model
Reflection
1 model
TypeSafe
1 model
ReMM SLERP 13B
1 model
Writer
1 model · US
The Palmyra models, targeted at regulated enterprise writing workflows.
XingChen
1 model