Model providers
Which models Oplema provides for Elara in each region, where your chats are processed and who hosts them.
Last updated
Elara
Elara Spark 1.0
- United States
- USDeepInfra, Sail Research through OpenRouter
- European Union
- EUMistral AI through Eden AI
- Asia
- US, GlobalDeepInfra, Sail Research, Cloudflare, DigitalOcean through OpenRouter
Elara Flame 1.0
- United States
- USAmazon Bedrock, Microsoft Azure through OpenRouter
- European Union
- EuropeMicrosoft Azure through OpenRouterDatabricks, Microsoft Azure through Eden AI
- Asia
- Global, US, EuropeAmazon Bedrock, Microsoft Azure, OpenAI through OpenRouter
OpenAI
GPT-6 Luna
- United States
- USMicrosoft Azure, Amazon Bedrock through OpenRouter
- European Union
- EuropeDatabricks through Eden AIMicrosoft Azure through OpenRouterMicrosoft Azure through Eden AI
- Asia
- Global, US, EuropeAmazon Bedrock, OpenAI, Microsoft Azure through OpenRouter
GPT-6.1 Sol
- United States
- USAmazon Bedrock, Microsoft Azure through OpenRouter
- European Union
- EuropeMicrosoft Azure through OpenRouterDatabricks, Microsoft Azure through Eden AI
- Asia
- Global, US, EuropeAmazon Bedrock, Microsoft Azure, OpenAI through OpenRouter
GPT-6 Astra
- United States
- USAmazon Bedrock, Microsoft Azure through OpenRouter
- European Union
- EUDatabricks through Eden AI
- Asia
- Global, USAmazon Bedrock, Microsoft Azure, OpenAI through OpenRouter
Anthropic
Claude Haiku 5.5
- United States
- USGoogle Vertex AI, Microsoft Azure, Amazon Bedrock through OpenRouter
- European Union
- EuropeDatabricks, Amazon Bedrock, Google Vertex AI through Eden AIGoogle Vertex AI, Amazon Bedrock through OpenRouter
- Asia
- Global, Europe, USGoogle Vertex AI, Anthropic, Microsoft Azure, Amazon Bedrock through OpenRouter
Claude Sonnet 5.5
- United States
- USAmazon Bedrock, Google Vertex AI, Microsoft Azure through OpenRouter
- European Union
- EuropeAmazon Bedrock, Databricks, Google Vertex AI through Eden AIAmazon Bedrock, Google Vertex AI through OpenRouter
- Asia
- Global, Europe, USGoogle Vertex AI, Amazon Bedrock, Microsoft Azure, Anthropic through OpenRouter
Claude Opus 5.5
- United States
- USAmazon Bedrock, Microsoft Azure, Google Vertex AI through OpenRouter
- European Union
- EuropeAmazon Bedrock, Databricks, Google Vertex AI through Eden AIAmazon Bedrock, Google Vertex AI through OpenRouter
- Asia
- Global, Europe, USAmazon Bedrock, Microsoft Azure, Google Vertex AI, Anthropic through OpenRouter
Claude Fable 5.1
- United States
- Not available
- European Union
- EUGoogle Vertex AI through Eden AI
- Asia
- GlobalMicrosoft Azure, Anthropic, Google Vertex AI through OpenRouter
DeepSeek
DeepSeek V4.1 Flash
- United States
- USDeepInfra, Sail Research through OpenRouter
- European Union
- EUMistral AI through Eden AI
- Asia
- US, GlobalDeepInfra, Sail Research, Cloudflare, DigitalOcean through OpenRouter
DeepSeek V4 Pro
- United States
- USDeepInfra, Sail Research through OpenRouter
- Asia
- US, GlobalDeepInfra, Nebius, CoreWeave, DigitalOcean, Cloudflare, Sail Research, Baseten through OpenRouter
Z.ai / GLM
GLM-5.3 Flash
- United States
- USDeepInfra, Sail Research, Fireworks through OpenRouter
- European Union
- EUGreenference, GreenPT, Mistral AI, TensorX through Eden AI
- Asia
- Global, USNEAR AI, CoreWeave, DigitalOcean, Makora, NextBit, Mistral AI, Nebius, Modal, Crusoe, Fireworks, FriendliAI, Baseten, DeepInfra, Sail Research through OpenRouter
GLM-5.3
- United States
- USDeepInfra, Sail Research, Baseten through OpenRouter
- European Union
- EuropeMistral AI through Eden AIMistral AI through OpenRouterGreenPT, TensorX through Eden AI
- Asia
- US, Global, EuropeDecart, Makora, DeepInfra, Cloudflare, Sail Research, CoreWeave, DigitalOcean, Mistral AI, Baseten, FriendliAI, Nebius, Crusoe, Modal, Fireworks through OpenRouter
xAI / Grok
Grok 4.7
- United States
- USxAI through OpenRouter
- European Union
- Not available
- Asia
- Global, USxAI, Amazon Bedrock through OpenRouter
Mistral
Mistral Small 4
- United States
- USMistral AI through OpenRouter
- European Union
- EuropeMistral AI through Eden AIMistral AI through OpenRouter
- Asia
- Global, Europe, USMistral AI through OpenRouter
Mistral Large 4
- United States
- Not available
- European Union
- EuropeMistral AI through Eden AIMistral AI through OpenRouter
- Asia
- Global, EuropeMistral AI through OpenRouter
Google Gemini
Gemini 3.5 Flash-Lite
- United States
- USGoogle Vertex AI through OpenRouter
- European Union
- EuropeGoogle Vertex AI through Eden AIGoogle Vertex AI through OpenRouter
- Asia
- Global, Europe, USGoogle Gemini API, Google Vertex AI through OpenRouter
Gemini 3.8 Flash
- United States
- USGoogle Vertex AI through OpenRouter
- European Union
- EUGoogle Vertex AI through Eden AI
- Asia
- Global, USGoogle Gemini API, Google Vertex AI through OpenRouter
Gemini 4 Argon
- United States
- USGoogle Vertex AI through OpenRouter
- European Union
- EUGoogle Vertex AI through Eden AI
- Asia
- Global, USGoogle Vertex AI, Google Gemini API through OpenRouter
MiniMax
MiniMax M3.1 Flash
- United States
- Not available
- Asia
- GlobalFireworks, DigitalOcean, Nebius, SambaNova through OpenRouter
Model prices
USD per million tokens by account region, as of November 13, 2026. These are the first listed host's prices, including the gateway's fee and Oplema's markup. If a later host answers instead, you pay the cost OpenRouter or Eden AI reports for that chat, with its gateway's fee and the same markup: at most 3x these prices, or 1.5 USD per million tokens where a price here is free. Prices can change.
Elara 2 models
Elara Spark 1.0
- United States
Input$0.4234
Cached input$0.0127
Cache writen/a
Output$1.2701
- European Union
Input$1.3926
Cached input$0.0325
Cache writen/a
Output$5.5704
- Asia
Input$0.4234
Cached input$0.0127
Cache writen/a
Output$1.2701
Elara Flame 1.0
- United States
Input$9.504
Cached input$0.4752
Cache write$11.88
Output$47.52
- European Union
Input$9.504
Cached input$0.4752
Cache write$11.88
Output$47.52
- Asia
Input$8.64
Cached input$0.432
Cache write$10.80
Output$43.20
Elara Flame 1.0, prompts above 272k tokens
- United States
Input$19.008
Cached input$0.9504
Cache write$23.76
Output$71.28
- European Union
Input$19.008
Cached input$0.9504
Cache write$23.76
Output$71.28
- Asia
Input$17.28
Cached input$0.864
Cache write$21.60
Output$64.80
OpenAI 3 models
GPT-6 Luna
- United States
Input$0.4752
Cached input$0.0475
Cache write$0.594
Output$2.376
- European Union
Input$0.422
Cached input$0.0422
Cache write$0.5276
Output$2.11
- Asia
Input$0.432
Cached input$0.0432
Cache write$0.54
Output$2.16
GPT-6 Luna, prompts above 272k tokens
- United States
Input$0.9504
Cached input$0.095
Cache write$1.188
Output$3.564
- European Union
Input$0.844
Cached input$0.0845
Cache write$1.0549
Output$3.1637
- Asia
Input$0.864
Cached input$0.0864
Cache write$1.08
Output$3.24
GPT-6.1 Sol
- United States
Input$9.504
Cached input$0.4752
Cache write$11.88
Output$47.52
- European Union
Input$9.504
Cached input$0.4752
Cache write$11.88
Output$47.52
- Asia
Input$8.64
Cached input$0.432
Cache write$10.80
Output$43.20
GPT-6.1 Sol, prompts above 272k tokens
- United States
Input$19.008
Cached input$0.9504
Cache write$23.76
Output$71.28
- European Union
Input$19.008
Cached input$0.9504
Cache write$23.76
Output$71.28
- Asia
Input$17.28
Cached input$0.864
Cache write$21.60
Output$64.80
GPT-6 Astra
- United States
Input$47.52
Cached input$4.752
Cache write$59.40
Output$237.60
- European Union
Input$42.20
Cached input$4.22
Cache write$52.75
Output$211
- Asia
Input$43.20
Cached input$4.32
Cache write$54
Output$216
GPT-6 Astra, long prompts
- United States
Input$95.04 above 272k tokens
Cached input$9.504 above 272k tokens
Cache write$118.80 above 272k tokens
Output$356.40 above 272k tokens
- European Union
Input$84.40 above 32k tokens
Cached input$8.44 above 32k tokens
Cache write$105.50 above 32k tokens
Output$316.50 above 32k tokens
- Asia
Input$86.40 above 272k tokens
Cached input$8.64 above 272k tokens
Cache write$108 above 272k tokens
Output$324 above 272k tokens
Anthropic 4 models
Claude Haiku 5.5
- United States
Input$0.4752
Cached input$0.0475
Cache write$0.594
Output$2.376
- European Union
Input$0.422
Cached input$0.0422
Cache write$0.5276
Output$2.11
- Asia
Input$0.432
Cached input$0.0432
Cache write$0.54
Output$2.16
Claude Haiku 5.5, prompts above 100k tokens
- United States
Input$2.376
Cached input$0.2376
Cache write$2.97
Output$11.88
- European Union
Input$2.11
Cached input$0.2109
Cache write$2.6376
Output$10.55
- Asia
Input$2.16
Cached input$0.216
Cache write$2.70
Output$10.80
Claude Sonnet 5.5
- United States
Input$9.504
Cached input$0.4752
Cache write$11.88
Output$47.52
- European Union
Input$9.284
Cached input$0.4642
Cache write$11.605
Output$46.42
- Asia
Input$8.64
Cached input$0.432
Cache write$10.80
Output$43.20
Claude Opus 5.5
- United States
Input$19.008
Cached input$0.9504
Cache write$23.76
Output$95.04
- European Union
Input$18.568
Cached input$0.9284
Cache write$23.21
Output$92.84
- Asia
Input$17.28
Cached input$0.864
Cache write$21.60
Output$86.40
Claude Fable 5.1
- United States
- Not available
- European Union
Input$46.42
Cached input$1.1605
Cache write$58.025
Output$232.10
- Asia
Input$43.20
Cached input$1.08
Cache write$54
Output$216
DeepSeek 2 models
DeepSeek V4.1 Flash
- United States
Input$0.4234
Cached input$0.0127
Cache writen/a
Output$1.2701
- European Union
Input$1.3926
Cached input$0.0325
Cache writen/a
Output$5.5704
- Asia
Input$0.4234
Cached input$0.0127
Cache writen/a
Output$1.2701
DeepSeek V4 Pro
- United States
Input$5.616
Cached input$0.432
Cache writen/a
Output$11.232
- European Union
Input$8.44
Cached input$2.11
Cache writen/a
Output$16.88
- Asia
Input$5.616
Cached input$0.432
Cache writen/a
Output$11.232
Z.ai / GLM 2 models
GLM-5.3 Flash
- United States
Input$0.648
Cached input$0.1296
Cache writen/a
Output$2.16
- European Union
Input$0.5486
Cached input$0.1266
Cache write$0
Output$1.8779
- Asia
Input$0.3456
Cached input$0.0518
Cache writen/a
Output$2.16
GLM-5.3
- United States
Input$2.43
Cached input$0.54
Cache writen/a
Output$10.80
- European Union
Input$6.4988
Cached input$0.6499
Cache writen/a
Output$20.425
- Asia
Input$2.4624
Cached input$0.4752
Cache writen/a
Output$8.1648
xAI / Grok 1 model
Grok 4.7
- United States
Input$9.504
Cached input$2.376
Cache writen/a
Output$28.512
- European Union
- Not available
- Asia
Input$8.64
Cached input$2.16
Cache writen/a
Output$25.92
Grok 4.7, prompts above 200k tokens
- United States
Input$19.008
Cached input$4.752
Cache writen/a
Output$57.024
- European Union
- Not available
- Asia
Input$17.28
Cached input$4.32
Cache writen/a
Output$51.84
Mistral 2 models
Mistral Small 4
- United States
Input$0.7128
Cached input$0.0713
Cache writen/a
Output$2.8512
- European Union
Input$0.6963
Cached input$0.0696
Cache writen/a
Output$2.7852
- Asia
Input$0.648
Cached input$0.0648
Cache writen/a
Output$2.592
Mistral Large 4
- United States
- Not available
- European Union
Input$3.2494
Cached input$0.3249
Cache writen/a
Output$9.7482
- Asia
Input$3.024
Cached input$0.3024
Cache writen/a
Output$9.072
Google Gemini 3 models
Gemini 3.5 Flash-Lite
- United States
Input$1.4256
Cached input$0.1426
Cache write$0.36
Output$11.88
- European Union
Input$1.3926
Cached input$0.1393
Cache write$0.3868
Output$11.605
- Asia
Input$1.296
Cached input$0.1296
Cache write$0.36
Output$10.80
Gemini 3.8 Flash
- United States
Input$3.564
Cached input$0.3564
Cache write$0.18
Output$17.82
- European Union
Input$3.4815
Cached input$0.3482
Cache write$0.1934
Output$17.407
- Asia
Input$3.24
Cached input$0.324
Cache write$0.18
Output$16.20
Gemini 4 Argon
- United States
Input$9.504
Cached input$0.4752
Cache write$0.81
Output$47.52
- European Union
Input$9.284
Cached input$0.4642
Cache write$0.8704
Output$46.42
- Asia
Input$8.64
Cached input$0.432
Cache write$0.81
Output$43.20
MiniMax 1 model
MiniMax M3.1 Flash
- United States
- Not available
- European Union
Input$1.266
Cached input$0.3165
Cache writen/a
Output$5.064
- Asia
Input$0.648
Cached input$0.0648
Cache writen/a
Output$2.592
Where your chats are processed
United States
Accounts in United States (US West) and United States (US East)
Chats are processed in the United States.
European Union
Accounts in European Union (Germany)
Models marked EU are processed in the EU. Models marked Europe are processed in the EU, the EEA, the UK or Switzerland.
Asia
Accounts in Asia (Singapore)
There's no location restriction, so each model's host can process your chats in any country.
Gateways and hosts
Elara sends each chat through OpenRouter, or for EU accounts OpenRouter or Eden AI, to a host that runs the model. Hosts are tried in the order listed. Each company's terms and data processing agreement cover how it handles your chats.
OpenRouter
OpenRouter, Inc., United States
US AI gateway that routes requests to many model providers through one API.
Eden AI
Eden AI SAS, France
French AI gateway that offers one API over many model providers.
Amazon Bedrock
Amazon Web Services, Inc., United States
Amazon Web Services' managed service for foundation models from several developers.
Anthropic
Anthropic, PBC, United States
US AI company that develops the Claude models.
Microsoft Azure
Microsoft Corporation, United States
Microsoft's cloud platform, which hosts OpenAI models through Microsoft Foundry.
Baseten
Baseten Labs, Inc., United States
US platform that serves AI models for inference.
Cloudflare
Cloudflare, Inc., United States
US cloud network that serves open-weight models through Workers AI.
CoreWeave
CoreWeave, Inc., United States
US GPU cloud that serves open-weight models for inference.
Crusoe
Crusoe Technologies LLC, United States
US AI cloud provider for open-weight model inference.
Databricks
Databricks, Inc., United States
US data and AI platform that serves hosted foundation models.
Decart
Decart.ai Inc., United States
US inference provider for open-weight models.
DeepInfra
Deep Infra Inc., United States
US inference provider for open-weight models.
DigitalOcean
DigitalOcean, LLC, United States
US cloud provider with serverless model inference.
Fireworks
Fireworks.ai, Inc., United States
US inference provider for open-weight models.
FriendliAI
FriendliAI Corp, United States
US inference provider for open-weight models.
Google Gemini API
Google LLC, United States
Google's developer API for the Gemini models.
Google Vertex AI
Google LLC, United States
Google Cloud's AI platform, which serves Gemini and partner models such as Claude.
Greenference
Greenference SAS, France
French inference provider for open-weight models, processing only in the EU.
GreenPT
GreenPT BV, Netherlands
Dutch inference provider for open-weight models, hosted in France and Finland.
Makora
A2 Labs, Inc., United States
US inference provider for open-weight models.
Mistral AI
Mistral AI, France
French AI company that develops the Mistral models and hosts other open-weight models.
Modal
Modal Labs, Inc., United States
US cloud platform that serves AI models for inference.
NEAR AI
Jasnah, Inc., United States
US inference provider that runs open-weight models in confidential computing enclaves.
Nebius
Nebius B.V., Netherlands
Dutch cloud provider for open-weight model inference.
NextBit
NEXTBIT 256, S.L., Spain
Spanish inference provider for open-weight models.
OpenAI
OpenAI, L.L.C., United States
US AI company that develops the GPT models.
Sail Research
Sail Research Co., United States
US inference provider for open-weight models.
SambaNova
SambaNova Systems, Inc., United States
US AI chip company that serves open-weight models.
TensorX
TensorX Ltd., Ireland
Irish inference provider for open-weight models, with GPUs in Dublin and Helsinki.
xAI
X.AI LLC, United States
US AI company that develops the Grok models.
How we handle your chats: Privacy policy, DPA section 9, Subprocessors, Elara