Legal
Models
Who processes each model you can pick in the platform and where, along with what it accepts, how much context it handles and its relative cost. The ones the platform itself uses by default are marked.
Where each provider processes data and under which safeguards: sub-processor register.
92 models from 13 authors, 31 of them processed in the EU.
92 of 95 modelsUsed by the platform
Alibaba Cloud (Tongyi Lab / Qwen)
1 model
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
qwen/qwen3-coderApache-2.0 | OpenRouter, Inc.Global | ImageAudioVideoPDF | 1M65.5K output | Jun 2025 | Jul 2025 |
Anthropic PBC
17 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
Used by the platformclaude-opus-5-5 | Anthropic PBCGlobal | ImageAudioVideoPDF | 1M128K output | Jun 2026 | Sep 2026 | ||
Used by the platformclaude-opus-5-5 | Google LLCEU | ImageAudioVideoPDF | 1M128K output | Jun 2026 | Sep 2026 | ||
anthropic/claude-opus-5 | OpenRouter, Inc.EU | ImageAudioVideoPDF | 1M128K output | Not published | Jul 2026 | ||
Used by the platformclaude-opus-5 | Anthropic PBCGlobal | ImageAudioVideoPDF | 1M128K output | Not published | Jul 2026 | ||
claude-opus-4-7 | Anthropic PBCGlobal | ImageAudioVideoPDF | 1M128K output | Jan 2026 | Apr 2026 | ||
claude-sonnet-4-6 | Anthropic PBCGlobal | ImageAudioVideoPDF | 1M64K output | Aug 2025 | Feb 2026 | ||
claude-opus-4-6 | Anthropic PBCGlobal | ImageAudioVideoPDF | 1M128K output | May 2025 | Feb 2026 | ||
claude-opus-4-5 | Anthropic PBCGlobal | ImageAudioVideoPDF | 200K64K output | May 2025 | Nov 2025 | ||
claude-haiku-4-5 | Anthropic PBCGlobal | ImageAudioVideoPDF | 200K64K output | Feb 2025 | Oct 2025 | ||
claude-sonnet-4-5 | Anthropic PBCGlobal | ImageAudioVideoPDF | 200K64K output | Jan 2025 | Sep 2025 | ||
anthropic/claude-sonnet-5 | OpenRouter, Inc.EU | ImageAudioVideoPDF | 1M128K output | Not published | Not published | ||
claude-haiku-5-5 | Anthropic PBCGlobal | ImageAudioVideoPDF | 1M128K output | Not published | Not published | ||
Used by the platformclaude-haiku-5-5 | Google LLCEU | ImageAudioVideoPDF | 1M128K output | Not published | Not published | ||
Used by the platformclaude-opus-4-8 | Anthropic PBCGlobal | ImageAudioVideoPDF | 1M128K output | Not published | Not published | ||
claude-sonnet-5 | Anthropic PBCGlobal | ImageAudioVideoPDF | 1M128K output | Not published | Not published | ||
claude-sonnet-5-5 | Anthropic PBCGlobal | ImageAudioVideoPDF | 1M128K output | Not published | Not published | ||
claude-sonnet-5-5 | Google LLCEU | ImageAudioVideoPDF | 1M128K output | Not published | Not published |
Beijing Zhipu Huazhang Technology Co., Ltd. (Z.ai)
4 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
z-ai/glm-5.3-flashMIT | OpenRouter, Inc.EUGlobal | ImageAudioVideoPDF | 1M131.1K output | Not published | Aug 2026 | ||
z-ai/glm-5.2MIT | OpenRouter, Inc.EUGlobal | ImageAudioVideoPDF | 1M262.1K output | Not published | Jun 2026 | ||
z-ai/glm-5.1MIT | OpenRouter, Inc.Global | ImageAudioVideoPDF | 202.8K131.1K output | Not published | Apr 2026 | ||
z-ai/glm-4.7MIT | OpenRouter, Inc.Global | ImageAudioVideoPDF | 202.8K131.1K output | Sep 2024 | Dec 2025 |
Google LLC
19 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
Used by the platformgemini-3.8-flash-tts | Google LLCGlobal | ImageAudioVideoPDF | Not published | Not published | Sep 2026 | ||
Used by the platformgemini-3.8-flash | Google LLCEUGlobal | ImageAudioVideoPDF | 1M65.5K output | Mar 2026 | Sep 2026 | ||
Used by the platformgemini-omni-1.1-flash | Google LLCGlobal | ImageAudioVideoPDF | Not published | Not published | Aug 2026 | ||
Used by the platformgemini-3.5-transcribe | Google LLCGlobal | ImageAudioVideoPDF | Not published | Not published | Aug 2026 | ||
gemini-3.7-flash | Google LLCEUGlobal | ImageAudioVideoPDF | 1M65.5K output | Mar 2026 | Aug 2026 | ||
Used by the platformgemini-3.5-flash-lite | Google LLCEUGlobal | ImageAudioVideoPDF | 1M65.5K output | Not published | Jul 2026 | ||
gemini-3.6-flash | Google LLCEUGlobal | ImageAudioVideoPDF | 1M65.5K output | Jan 2025 | Jul 2026 | ||
gemini-3.5-flash | Google LLCEUGlobal | ImageAudioVideoPDF | 1M65.5K output | Jan 2025 | May 2026 | ||
gemini-3.1-flash-lite | Google LLCEUGlobal | ImageAudioVideoPDF | 1M65.5K output | Jan 2025 | May 2026 | ||
Used by the platformgemini-embedding-2 | Google LLCEU | ImageAudioVideoPDF | 8.2K | Not published | Apr 2026 | ||
gemini-3.1-pro-preview | Google LLCGlobal | ImageAudioVideoPDF | 1M65.5K output | Jan 2025 | Feb 2026 | ||
gemini-3.1-pro-preview-customtools | Google LLCGlobal | ImageAudioVideoPDF | 1M65.5K output | Jan 2025 | Feb 2026 | ||
gemini-3-flash-preview | Google LLCGlobal | ImageAudioVideoPDF | 1M65.5K output | Jan 2025 | Dec 2025 | ||
gemini-2.5-flash | Google LLCEUGlobal | ImageAudioVideoPDF | 1M65.5K output | Jan 2025 | Jun 2025 | ||
gemini-2.5-pro | Google LLCEUGlobal | ImageAudioVideoPDF | 1M65.5K output | Jan 2025 | Jun 2025 | ||
Used by the platformgemini-2.5-flash-tts | Google LLCEU | ImageAudioVideoPDF | Not published | Not published | Not published | ||
Used by the platformgemini-3.1-flash-image | Google LLCEUGlobal | ImageAudioVideoPDF | Not published | Jan 2025 | Not published | ||
lyria-3-clip-preview | Google LLCGlobal | ImageAudioVideoPDF | Not published | Not published | Not published | ||
Used by the platformlyria-3.5 | Google LLCGlobal | ImageAudioVideoPDF | Not published | Not published | Not published |
Hangzhou DeepSeek AI Basic Technology Research Co., Ltd.
6 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
deepseek/deepseek-v4.1-flashMIT | OpenRouter, Inc.Global | ImageAudioVideoPDF | 1M384K output | Not published | Sep 2026 | ||
deepseek/deepseek-v4-flash-vision-expMIT | OpenRouter, Inc.Global | ImageAudioVideoPDF | 1M384K output | Not published | Aug 2026 | ||
deepseek/deepseek-v4-pro-0813MIT | OpenRouter, Inc.Global | ImageAudioVideoPDF | 1M393.2K output | Not published | Aug 2026 | ||
deepseek/deepseek-v4-flashMIT | OpenRouter, Inc.Global | ImageAudioVideoPDF | 1M393.2K output | Not published | Apr 2026 | ||
deepseek/deepseek-v4-proMIT | OpenRouter, Inc.Global | ImageAudioVideoPDF | 1M384K output | Not published | Apr 2026 | ||
deepseek/deepseek-v3.2MIT | OpenRouter, Inc.Global | ImageAudioVideoPDF | 131.1K64K output | Sep 2025 | Dec 2025 |
Meta Platforms, Inc.
2 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
meta-llama/llama-4-maverickLlama 4 Community License | OpenRouter, Inc.Global | ImageAudioVideoPDF | 1M16.4K output | Aug 2024 | Apr 2025 | ||
meta-llama/llama-4-scoutLlama 4 Community License | OpenRouter, Inc.Global | ImageAudioVideoPDF | 10M16.4K output | Aug 2024 | Apr 2025 |
Microsoft
3 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
Used by the platformmicrosoft-decision-1 | Microsoft CorporationEU | ImageAudioVideoPDF | 32.8K | Not published | Oct 2026 | ||
Used by the platformfast-transcription | Microsoft CorporationEU | ImageAudioVideoPDF | Not published | Not published | Not published | ||
Used by the platformneural-tts | Microsoft CorporationEU | ImageAudioVideoPDF | Not published | Not published | Not published |
MiniMax
2 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
minimax/minimax-m3 | OpenRouter, Inc.Global | ImageAudioVideoPDF | 1M512K output | Not published | Jun 2026 | ||
minimax/minimax-m2.5Modified-MIT | OpenRouter, Inc.Global | ImageAudioVideoPDF | 204.8K196.6K output | Not published | Feb 2026 |
Mistral AI
3 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
codestral-2 | Google LLCEU | ImageAudioVideoPDF | 128K | Not published | Aug 2025 | ||
mistral-medium-3 | Google LLCEU | ImageAudioVideoPDF | 128K | Not published | May 2025 | ||
mistral-small-2503Apache-2.0 | Google LLCEU | ImageAudioVideoPDF | 128K | Not published | Mar 2025 |
Moonshot AI (Beijing Moonshot AI Technology Co., Ltd.)
2 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
moonshotai/kimi-k3 | OpenRouter, Inc.Global | ImageAudioVideoPDF | 1M1M output | Not published | Jul 2026 | ||
moonshotai/kimi-k2-thinkingModified MIT License | OpenRouter, Inc.Global | ImageAudioVideoPDF | 262.1K262.1K output | Not published | Nov 2025 |
OpenAI, L.L.C.
27 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
Used by the platformgpt-6.1-sol | Microsoft CorporationEU | ImageAudioVideoPDF | 1.1M128K output | Apr 2026 | Sep 2026 | ||
Used by the platformgpt-6.1-sol | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 1.1M128K output | Apr 2026 | Sep 2026 | ||
Used by the platformgpt-5.6-luna | Microsoft CorporationEU | ImageAudioVideoPDF | 1.1M128K output | Feb 2026 | Jul 2026 | ||
Used by the platformgpt-5.6-luna | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 1.1M128K output | Feb 2026 | Jul 2026 | ||
gpt-5.6-sol | Microsoft CorporationEU | ImageAudioVideoPDF | 1.1M128K output | Feb 2026 | Jul 2026 | ||
gpt-5.6-sol | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 1.1M128K output | Feb 2026 | Jul 2026 | ||
gpt-5.6-terra | Microsoft CorporationEU | ImageAudioVideoPDF | 1.1M128K output | Feb 2026 | Jul 2026 | ||
gpt-5.6-terra | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 1.1M128K output | Feb 2026 | Jul 2026 | ||
Used by the platformgpt-realtime-whisper | OpenAI, L.L.C.Global | ImageAudioVideoPDF | Not published | Not published | May 2026 | ||
gpt-5.5 | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 1M128K output | Dec 2025 | Apr 2026 | ||
gpt-5.4-mini | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 400K128K output | Aug 2025 | Mar 2026 | ||
gpt-5.4 | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 1M128K output | Aug 2025 | Mar 2026 | ||
gpt-5.2 | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 400K128K output | Aug 2025 | Dec 2025 | ||
gpt-5.1 | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 272K128K output | Sep 2024 | Nov 2025 | ||
gpt-5 | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 272K128K output | Sep 2024 | Aug 2025 | ||
Used by the platformopenai/gpt-oss-120bApache-2.0 | Groq, Inc.Global | ImageAudioVideoPDF | 131.1K131.1K output | Jun 2024 | Aug 2025 | ||
gpt-4.1 | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 1M32.8K output | Jun 2024 | Apr 2025 | ||
Used by the platformgpt-4o-mini-transcribe | OpenAI, L.L.C.Global | ImageAudioVideoPDF | Not published | Not published | Mar 2025 | ||
Used by the platformgpt-4o-mini-tts | OpenAI, L.L.C.Global | ImageAudioVideoPDF | Not published | Not published | Mar 2025 | ||
Used by the platformgpt-4o-transcribe | Microsoft CorporationEU | ImageAudioVideoPDF | Not published | Not published | Mar 2025 | ||
gpt-4o-mini | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 128K16.4K output | Oct 2023 | Jul 2024 | ||
gpt-6-astra | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 1M128K output | Apr 2026 | Not published | ||
Used by the platformgpt-6-luna | Microsoft CorporationEU | ImageAudioVideoPDF | 1.1M128K output | May 2026 | Not published | ||
Used by the platformgpt-6-luna | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 1.1M128K output | May 2026 | Not published | ||
Used by the platformgpt-6-sol | Microsoft CorporationEU | ImageAudioVideoPDF | 1.1M128K output | Apr 2026 | Not published | ||
Used by the platformgpt-6-sol | OpenAI, L.L.C.Global | ImageAudioVideoPDF | 1.1M128K output | Apr 2026 | Not published | ||
openai/gpt-oss-120b-maasApache-2.0 | Google LLCGlobal | ImageAudioVideoPDF | 131.1K131.1K output | Not published | Not published |
X.AI LLC
5 models
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
grok-4.7 | X.AI LLCUS | ImageAudioVideoPDF | 500K | May 2026 | Sep 2026 | ||
grok-4.6 | X.AI LLCUS | ImageAudioVideoPDF | 500K | Feb 2026 | Aug 2026 | ||
Used by the platformgrok-4.5 | X.AI LLCUS | ImageAudioVideoPDF | 500K | Not published | Jul 2026 | ||
Used by the platformgrok-4.3 | X.AI LLCGlobal | ImageAudioVideoPDF | 1M | Dec 2025 | Apr 2026 | ||
grok-4.20-reasoning | X.AI LLCUS | ImageAudioVideoPDF | 2M30K output | Not published | Not published |
Xiaomi
1 model
| Model | Processed by | Inputs | Reasons | Context | Data until | Released | Cost |
|---|---|---|---|---|---|---|---|
xiaomi/mimo-v2.5-proMIT | OpenRouter, Inc.Global | ImageAudioVideoPDF | 1M131.1K output | Not published | Apr 2026 |
How we choose them
- Every model the platform calls is listed here, with who processes it and where.
- We add a model on a second route only when that route reaches a model we could not reach otherwise or processes it in a region a customer needs.
- What we say about each provider's data handling comes from its own data processing agreement and sub-processor pages.
- No request reaches a host run by a Chinese lab or its offshore arm. Open weight models from those labs are served by other hosts.
- A workspace held to EU processing reaches only models processed in the EU.
Cost is each model's output price in three steps. Your usage is billed in credits according to your plan.