the board members
Providers
Two kinds of sellers set the market: labs that make models, and hosts that serve open-weight models on their own hardware. Prices for the same weights can differ meaningfully between hosts — and so can where your prompts go. Filter by jurisdiction, or by whether a provider can run with zero prompt retention.
Aggregators
one API across closed and open modelsstandardcompute
Norway · Other HQ
Closed and open frontier models behind one OpenAI-compatible API, sold as flat monthly plans instead of per-token metering — built by the team behind this site.
Flat plans from $19/mo · effectively ~30% under list at sustained volume
OpenRouter
United States · US HQ
Routes one API across hundreds of closed and open models from every major lab, at the providers' own list prices.
300+ models · pass-through list pricing
Model labs
they train the models they sellOpenAI
28 modelsUnited States · US HQ · runs in US, EU, Other
Maker of the GPT family. The default choice for many production apps, with a wide spread from nano-tier workhorses to frontier reasoning models.
From $0.20/1M input · GPT-5.6 Luna
Anthropic
10 modelsUnited States · US HQ · runs in US
Maker of Claude. Strong on long-context work, coding agents, and careful instruction following, with aggressive prompt-caching discounts.
From $1/1M input · Claude Haiku 4.5
United States · US HQ
Gemini API models with huge context windows and some of the best price-to-capability ratios in the mid tier.
From $0.25/1M input · Gemini 3.1 Flash-Lite
DeepSeek
3 modelsChina · China HQ · runs in China
Open-weight frontier challenger known for rock-bottom token prices and strong reasoning performance per dollar.
From $0.44/1M input · DeepSeek V4 Flash
xAI
7 modelsUnited States · US HQ
Maker of Grok. Fast-shipping frontier lab with real-time X data integration and competitive flagship pricing.
From $1/1M input · Grok Build 0.1
Mistral
7 modelsFrance · EU HQ · runs in EU, US
European lab shipping efficient open and commercial models, popular where data residency and price both matter.
From $0.10/1M input · Ministral 3 3B
Moonshot AI
4 modelsChina · China HQ
Maker of the Kimi models — open-weight frontier contenders with standout agentic coding results at challenger prices.
From $0.95/1M input · Kimi K2.7 Code
Alibaba Qwen
8 modelsChina · China HQ · runs in Other, EU, US
The Qwen family spans tiny budget models to frontier-class flagships, served via Alibaba Cloud Model Studio.
From $0.05/1M input · Qwen Flash
Z.ai
13 modelsChina · China HQ · runs in Other
Maker of the GLM series, best known for coding-oriented models and unusually cheap coding subscription plans.
From $0.07/1M input · GLM-4.7-FlashX
MiniMax
6 modelsChina · China HQ
Shanghai-based lab with long-context open-weight models priced for high-volume workloads.
From $0.30/1M input · MiniMax M3
Cohere
7 modelsCanada · Other HQ
Enterprise-focused lab behind the Command family, with strengths in RAG, tool use, and multilingual work.
From $0.04/1M input · Command R7B
Perplexity
4 modelsUnited States · US HQ
Sonar API models that pair generation with built-in web search — priced per token plus search fees.
From $1/1M input · Sonar
Amazon
6 modelsUnited States · US HQ · runs in US, EU, Other
Amazon's Nova family on Bedrock — priced aggressively for high-volume enterprise workloads inside AWS.
From $0.04/1M input · Nova Micro
AI21 Labs
2 modelsIsrael · Other HQ
Maker of the Jamba models — hybrid SSM-transformer architecture with long context at mid-tier prices.
From $0.20/1M input · Jamba Mini
Inference hosts
they serve open-weight modelsMicrosoft Azure
4 modelsUnited States · US HQ · runs in US, EU, Other
OpenAI's models served on Azure AI Foundry — same rates as OpenAI direct, with Azure networking, compliance, and billing.
From $0.20/1M input · GPT-5.6 Luna
Fireworks AI
13 modelsUnited States · US HQ · runs in US, EU, Other
Fast serverless inference for the popular open-weight models, with a production focus and per-token pricing.
From $0.05/1M input · NVIDIA Nemotron 3.5 Lightning 30B
DeepInfra
13 modelsUnited States · US HQ · runs in US
Consistently among the cheapest hosts for open-weight models, with simple per-token serverless pricing.
From $0.08/1M input · DeepSeek-V4-Flash-0731
Cerebras
2 modelsUnited States · US HQ · runs in US
Wafer-scale hardware serving open models at extreme token throughput — speed is the headline, price stays competitive.
From $0.35/1M input · GPT OSS 120B
Nebius
12 modelsNetherlands · EU HQ · runs in EU, US, Other
European AI cloud with serverless open-model endpoints and budget-friendly token rates.
From $0.14/1M input · DeepSeek-V4-Flash
Novita AI
11 modelsHQ not disclosed
Broad low-cost catalog of open-weight models on serverless endpoints.
From $0.05/1M input · OpenAI GPT OSS 120B
Groq
7 modelsUnited States · US HQ · runs in US, EU, Other
Custom LPU hardware serving open models at very high speed. Often the cheapest way to run Llama-class models fast.
From $0.03/1M input · Llama Prompt Guard 2 22M
Together AI
16 modelsUnited States · US HQ · runs in US, EU
Broad catalog of open-weight models on serverless endpoints — a one-stop shop for Llama, DeepSeek, and Qwen inference.
From $0.05/1M input · GPT-OSS 20B