chatstream.net Query Parameter Specification
chatstream.net is a service operated by Qualiteg Inc. that lets you try out the latest publicly available LLMs from around the world as soon as they are released.
You can open a particular LLM for an in-depth chat, open several LLMs and use them together, or have LLMs compare their outputs side by side—giving you a real sense of what LLMs can do.
For example, opening a URL in a PC browser can launch four LLMs at once so you can chat with all of them simultaneously. URL parameters are what let you control chat behavior to your liking in this way.
URL Parameters
The behavior of chatstream.net can be controlled to some extent through URL parameters.
URL parameters work like this: after https://chatstream.net you append ? — for example https://chatstream.net?model_id=openai_gpt_4o_mini — and following the ? you specify a query string like key=value to control chatstream.net according to your preferences.
Parameter List
[ws_name]
Specifies the name of the working set to be selected automatically.
ws_name="chat_app"
A working set is a unit of work you can switch between using the buttons shown on the left edge of the screen on a PC (bottom edge on a smartphone). Currently, "chat_app" and "chat_app_en" can be specified.
[mult]
A query parameter representing multi_topic_mode.
mult=1
Specifying mult=1 enables multi-topic mode, which lets you open multiple LLM chats at the same time when using a PC.
(Examples)
mult=1: multi-topic mode
mult=0: single-topic mode
mult unspecified: default setting or the user's saved preference
[ontp]
A query parameter representing open_new_topic.
Opens the chat as a new topic.
ontp=1
(Examples)
ontp=1: automatically opens a new topic
ontp=0: (follows the default)
ontp unspecified: (follows the default behavior)
[model_id]
Specifies the model_id you want to open automatically.
Specifying multiple values opens multiple models.
model_id=openai_gpt_4o_mini&model_id=rakuten__rakuten_ai_7b_chat
List of model_ids (note: some models are available only in the enterprise edition)
| Model name | Model display name | model_id |
|---|---|---|
| llama3.1 | Meta-Llama-3.1-8B | meta_llama_3_1_8b_instruct |
| node (mistral_nemo for chatstream.net) | Mistral-Nemo-Instruct-2407 | mistral_nemo_instruct_2407 |
| node:0(default) | calm3-22b-chat | calm3_22b_chat |
| node (elyza8B for chatstream.net) | Llama-3-ELYZA-JP-8B | llama_3_elyza_jp_8b |
| node:0(default) | RakutenAI-7B-chat | rakuten__rakuten_ai_7b_chat |
| node (GPT4o_mini for chatstream.net) | OpenAI GPT-4o mini | openai_gpt_4o_mini |
| node (GPT3.5_newtech for chatstream.net) | OpenAI GPT-3.5 | openai_gpt_3_5_175b |
| node (Claude3Haiku for chatstream.net) | Anthropic Claude3-Haiku | anthropic_claude3_0_haiku |
| node (Gemini1.5Flash for chatstream.net) | Google Gemini1.5 Flash | google_gemini1_5_flash |
| Anthropic Claude 3.5 Sonnet | Anthropic Claude 3.5 Sonnet | anthropic_claude_3_5_sonnet |
| Google Gemini1.5 Pro | Google Gemini1.5 Pro | google_gemini1_5_pro |
| OpenAI gpt-4o | OpenAI GPT4o | openai_gpt4o |
Notes
- To make sure you open multiple models, explicitly specify mult=1.
- To make sure a topic opens automatically, explicitly specify ontp=1.
[isync]
A query parameter representing input_sync.
Lets you synchronize your input across multiple LLMs.
isync=1
isync=1: input synchronization enabled
isync=0: no input synchronization
isync unspecified: follows the default setting
[noip]
noip is a query parameter representing no_iphone.
Disables the iPhone-specific rendering mode.
