LLM prices per 1M tokens

264 models · 161 price changes

LLM API providers

ProviderModels
Model labs
OpenAI

GPT models through the OpenAI API.

61
Google Gemini API

Gemini and Gemma through the Gemini API (Google AI Studio).

33
Qwen (Alibaba)

Qwen models through Alibaba Cloud Model Studio, and open weights on hosts.

43
Mistral AI

Mistral models through La Plateforme.

24
xAI

Grok models through the xAI API.

7
Z.ai

GLM models through the Z.ai API.

17
Anthropic

Claude models through the Claude API.

15
Cohere

Command models through the Cohere API.

11
Meta

Llama (open weights, served by hosts) and Muse models.

16
DeepSeek

DeepSeek models through the DeepSeek API.

14
Amazon Nova

Amazon Nova models, sold through Bedrock.

8
Perplexity

Sonar search models, the Search API and embeddings through the Perplexity API.

7
Moonshot AI

Kimi models through the Moonshot platform.

8
NVIDIA

Nemotron open-weight models (served by hosts) and NIM microservices under an NVIDIA AI Enterprise licence.

6
Cloud platforms
Azure AI Foundry

OpenAI, Claude and other models, billed through Azure.

60
Google Vertex AI

Gemini, Claude and open models, billed through Google Cloud.

35
AWS Bedrock

Claude, Llama, Mistral, Nova and more, billed through AWS.

43
Claude Platform on AWS

Anthropic's Claude Platform, billed through AWS.

12
Databricks Mosaic AI Gateway + Foundation Model APIs

Sells 8 of the models tracked here.

8
Snowflake Cortex AI (AI_COMPLETE)

Sells 7 of the models tracked here.

7
OCI Generative AI

Sells 4 of the models tracked here.

4
IBM watsonx.ai

Sells 3 of the models tracked here.

3
Inference hosts
DeepInfra

Serverless open-weight models.

57
SiliconFlow

Serverless open-weight models.

38
Parasail

Serverless open-weight models.

33
Venice

Private inference for open-weight models.

32
Novita AI

Serverless open-weight models.

30
CoreWeave

GPU cloud with serverless inference.

20
Phala

Sells 18 of the models tracked here, on 19 routes.

18
Baseten

Model APIs and dedicated inference.

12
AtlasCloud

Sells 18 of the models tracked here.

18
Nebius

Open-weight models on Nebius Token Factory.

17
GMICloud

Sells 17 of the models tracked here.

17
Cloudflare Workers AI

Open-weight models on Cloudflare's network.

16
DigitalOcean

Serverless inference on the Gradient platform.

16
Wafer

Sells 11 of the models tracked here, on 21 routes.

11
StreamLake

Sells 16 of the models tracked here.

16
Fireworks AI

Serverless open-weight models.

9
PrimeIntellect

Sells 13 of the models tracked here.

13
Baidu

Sells 12 of the models tracked here, on 13 routes.

12
Together AI

Serverless open-weight models.

12
Relace

Sells 11 of the models tracked here, on 13 routes.

11
Sail Research

Sells 8 of the models tracked here, on 15 routes.

8
DekaLLM

Sells 11 of the models tracked here.

11
InferenceNet

Sells 6 of the models tracked here, on 16 routes.

6
Crusoe

Open-weight models on Crusoe Cloud.

9
Modal

Serverless GPU compute.

9
Darkbloom

Sells 9 of the models tracked here.

9
Morph

Sells 8 of the models tracked here.

8
AkashML

Sells 8 of the models tracked here.

8
Chutes

Decentralized serverless inference.

7
Makora

Sells 7 of the models tracked here.

7
Reka

Sells 7 of the models tracked here.

7
NextBit

Sells 7 of the models tracked here.

7
Io Net

Sells 6 of the models tracked here.

6
Inceptron

Sells 6 of the models tracked here.

6
Mancer 2

Sells 6 of the models tracked here.

6
Groq

Open-weight models on Groq LPUs.

5
Friendli

Serverless open-weight models.

5
SambaNova

Open-weight models on SambaNova RDUs.

4
Near AI

Sells 4 of the models tracked here.

4
OpenInference

Sells 3 of the models tracked here.

3
Decart

Sells 3 of the models tracked here.

3
ModelRun

Sells 3 of the models tracked here.

3
Particle AI

Sells 2 of the models tracked here, on 3 routes.

2
Cerebras

Open-weight models on wafer-scale chips.

2
Ionstream

Sells 2 of the models tracked here.

2
Tenstorrent

Sells 2 of the models tracked here.

2
PrimaLabs

Sells 2 of the models tracked here.

2
ParetoAlpha

Sells 1 of the models tracked here.

1
Ambient

Sells 1 of the models tracked here.

1
StepFun

Sells 1 of the models tracked here.

1

Official price lists

Companies whose model access is priced on their own pages but not in the daily catalog: embeddings, search APIs, credits, GPU hours and licences, read by hand with a link to each page.

ProviderPrices
Voyage AI by MongoDB
Model lab
21
Cursor
Model lab
3
Hugging Face
Inference host
8
Elastic Inference Service
Cloud platform
2

Found, not tracked yet

Labs selling token-priced models on OpenRouter that this site does not track yet. Unverified: the lab's own pricing page has not been read. 174 more AI companies from the companies.fru.dev registry are checked every morning for priced models.

LabModels
Tencent
Unverified; Hy4, Hy4 preview
8
AionLabs
Unverified; Aion 3.5 Mini, Aion 3.5
6
ByteDance Seed
Unverified; Seed 2.1 Turbo, Seed-2.0-Code
6
MiniMax
Unverified; MiniMax M3.1 Flash, MiniMax M3
5
Sentence Transformers
Unverified; paraphrase-MiniLM-L6-v2, all-MiniLM-L12-v2
5
VoyageAI by MongoDB
Unverified; voyage-code-4, voyage-multimodal-3.5
5
VoyageAI by MongoDB
Unverified; voyage-code-4, voyage-multimodal-3.5
5
Xiaomi
Unverified; MiMo-V2.6-Pro-UltraSpeed, MiMo-V2.6-Flash
5
Sakana
Unverified; Fugu Ultra v2, Fugu Max
4
inclusionAI
Unverified; Ling 3.1 Flash, Ling 3.0 Flash VL
4
BAAI
Unverified; bge-base-en-v1.5, bge-large-en-v1.5
3
Inception
Unverified; Mercury 3, Mercury 2.5
3
Intfloat
Unverified; E5-Large-v2, E5-Base-v2
3
LiquidAI
Unverified; LFM3-1.7B, LFM2.5-24B-A2B
3
StepFun
Unverified; Step 5 Preview, Step 3.7 Flash
3
TheDrummer
Unverified; Cydonia 24B V4.1, Skyfall 36B V2
3
Inference.net
Unverified; Schematron V2 Turbo, Schematron V2 Small
2
Moondream
Unverified; Moondream 3 Preview, Moondream 3.1
2
Morph
Unverified; Morph V3 Large, Morph V3 Fast
2
Nex AGI
Unverified; Nex-N2.5-Mini, Nex-N2.5-Pro
2
Perceptron
Unverified; Perceptron Mk1.5, Perceptron Mk1
2
Poolside
Unverified; Laguna S 2.1, Laguna XS 2.1
2
PrismML
Unverified; Ternary Bonsai 2 9B, Ternary Bonsai 2 27B
2
Relace
Unverified; Relace Search, Relace Apply 3
2
Sao10K
Unverified; Llama 3.1 Euryale 70B v2.2, Llama 3 8B Lunaris
2
ScaleDown
Unverified; Summarize, Extract
2
Thenlper
Unverified; GTE-Base, GTE-Large
2
Thinking Machines
Unverified; Inkling Small, Inkling
2
Upstage
Unverified; Solar Mini 4, Solar Pro 4
2
rekaai
Unverified; Reka Edge, Reka Flash 3
2
unbiased
Unverified; Pareto 26.10, Pareto 26.9
2
Arcee AI
Unverified; Trinity Large Thinking
1
Baidu
Unverified; ERNIE 4.5 VL 424B A47B
1
ByteDance
Unverified; UI-TARS 7B
1
Elorian
Unverified; Glimpse
1
Fireworks
Unverified; Ember-1
1
IBM
Unverified; Granite 4.2 8B
1
Kwaipilot
Unverified; KAT-Coder-Pro V2.5
1
Mancer
Unverified; Weaver (alpha)
1
Meituan
Unverified; LongCat 2.0
1
Microsoft
Unverified; Phi 4
1
Nous
Unverified; Hermes 4 405B
1
Particle AI
Unverified; Particle-1
1
Reflection
Unverified; Beam 1.0 501B A23B
1
Ricursive
Unverified; Ripple 65B Preview
1
Venice
Unverified; Uncensored
1
Writer
Unverified; Palmyra X5
1
anthracite-org
Unverified; Magnum v4 72B
1
gryphe
Unverified; MythoMax 13B
1
undi95
Unverified; ReMM SLERP 13B
1