Unofficial site — not affiliated with or endorsed by OpenRouter or any model maker. Chat answers are AI-generated and may be wrong.
Use any listed AI model now
331 AI models, one balance, pay for what you use
Pick one of 331 text models and chat with one credit balance: each reply costs credits by its real token cost (at least 1 credit, about $0.01), and a new account gets 5 free credits once, valid for 7 days. The 474 models we track are listed below with their prices.
- models to use now
- 331
- models listed
- 474
- last synced
- 2026-11-13
One balance
Chat with any of them
474 models
Model directory
Showing 12 of 474 models
- Google: Gemini 4 Carbongoogle/gemini-4-carbonText
- Maker
- Input / 1M
- $2.00$0. per input image
- Output / 1M
- $10
- Per reply
- ≈ 3 credits
Use hereChat - Thinking Machines: Inkling 2 Smallthinkingmachines/inkling-2-smallText
- Maker
- Thinking Machines
- Input / 1M
- $0.30
- Output / 1M
- $1.00
- Per reply
- ≈ 1 credit
Use hereChat - Anthropic: Claude Fable 5.2anthropic/claude-fable-5.2Text
- Maker
- Anthropic
- Input / 1M
- $8.00
- Output / 1M
- $40
- Per reply
- ≈ 9 credits
Use hereChat - DeepSeek: DeepSeek V4.1 Flash 1017deepseek/deepseek-v4.1-flash-1017Text
- Maker
- DeepSeek
- Input / 1M
- $0.04
- Output / 1M
- $0.50
- Per reply
- ≈ 1 credit
Use hereChat - OpenAI: GPT-6 Lunaopenai/gpt-6-lunaText
- Maker
- OpenAI
- Input / 1M
- $0.10
- Output / 1M
- $0.50
- Per reply
- ≈ 1 credit
Use hereChat - Google: Gemini 4 Hydrogengoogle/gemini-4-hydrogenText
- Maker
- Input / 1M
- $1.50$0. per input image
- Output / 1M
- $7.50
- Per reply
- ≈ 2 credits
Use hereChat - Baidu: ERNIE-Light-DT-1.0baidu/ernie-light-dt-1.0-intlText
- Maker
- Baidu
- Input / 1M
- $1.13
- Output / 1M
- $3.94
- Per reply
- ≈ 1 credit
Use hereChat - Nex AGI: Nex-N2.6nex-agi/nex-n2.6Text
- Maker
- Nex AGI
- Input / 1M
- $0.12
- Output / 1M
- $0.40
- Per reply
- ≈ 1 credit
Use hereChat - Meta: Muse Embermeta/muse-emberText
- Maker
- Meta
- Input / 1M
- $0.10
- Output / 1M
- $0.50
- Per reply
- ≈ 1 credit
Use hereChat - LiquidAI: LFM3-1.7Bliquid/lfm-3-1.7bText
- Maker
- Liquid AI
- Input / 1M
- $0.08
- Output / 1M
- $0.20
- Per reply
- ≈ 1 credit
Use hereChat - KT: Mi:dm K 3.0 Prokt/midm-k-3.0-proText
- Maker
- KT
- Input / 1M
- $0.20
- Output / 1M
- $0.80
- Per reply
- ≈ 1 credit
Use hereChat - Ricursive: Ripple 65B Previewricursiveai/ripple-65b-previewText
- Maker
- Ricursive
- Input / 1M
- $0.10
- Output / 1M
- $0.30
- Per reply
- ≈ 1 credit
Use hereChat
How this list and the prices are made
Rows come from OpenRouter's public model list (https://openrouter.ai/api/v1/models), fetched once a day and checked before it is stored (last sync: November 13, 2026). Input / output prices are OpenRouter's listed prices in US dollars per 1 million tokens. A text model can be used here only if OpenRouter offers it on a host that does not train on prompts (checked daily with this site's privacy settings; every request also asks for such hosts only); image, video and audio models are listed but not available here yet. "Per reply" is an estimate for a typical reply (about 300 words in, 500 words out); each real reply is billed by its own token cost.
FAQ
Questions
How much does a reply cost?
In proportion to its real token cost: the model's price for what you sent (your message plus the earlier messages sent along) and what it answered, with our service fee included, is converted into credits and rounded up, at least 1 credit; a credit is $0.01 in the $5 pack (the $10 and $20 packs add bonus credits, and monthly and yearly plans include more credits per dollar, down to $0.0070 on a yearly plan). Before the call the most a reply can cost is held; after it the difference is returned. The list shows an estimate per model for a typical reply.
Is there anything free?
A new account gets 5 free credits once, valid for 7 days. They work on every model in the chat, like any other credits; a reply on a pricier model can need more credits than that.
Which models can I use?
331 text models right now. A model is usable when OpenRouter offers it on a host that does not train on prompts, it has a fixed token price with no extra per-request or per-search fee, and it costs at most $20 per million input tokens and $80 per million output tokens. Image, video and audio models are listed but not available here yet.
Are my conversations stored?
Not on this site. The conversation is kept only in your browser; the server stores the model, credits, token counts and cost of each reply, not the text.
Where does the model list come from?
From OpenRouter's public model list (https://openrouter.ai/api/v1/models). The site fetches it once a day; the copy shown here was synced on November 13, 2026 and lists 474 models.