MediaWiki - Recent changes [en]

AI on Demand: Token streaming

Sst-la1 — Mon, 09 Nov 2026 15:36:28 GMT

Usage - Inference with token-stream

← Older revision		Revision as of 16:36, 9 November 2026
Line 16:		Line 16:

	<syntaxhighlight lang=bash>		<syntaxhighlight lang=bash>
			#Persönlichen Key setzen
			STONEY_KEY='sk-...'

	#Modell auswählen		#Modell auswählen
	MODEL='zai-org/GLM-5.3-Flash'		MODEL='zai-org/GLM-5.3-Flash'

AI on Demand: Token streaming

Sst-djo — Mon, 09 Nov 2026 15:30:04 GMT

Usage - Inference with token-stream

← Older revision		Revision as of 16:30, 9 November 2026
(One intermediate revision by the same user not shown)
Line 1:		Line 1:
	= Overview =		= Overview =
	This pages describes some useful commands on how to ~~test~~ token streaming.		This pages describes some useful commands on how to use token streaming.

	= Usage =		= Usage =
Line 12:		Line 12:

	Chat platforms like LibreChat include native mechanisms for handling token streaming or parsing the stream.		Chat platforms like LibreChat include native mechanisms for handling token streaming or parsing the stream.

			The stream: true flag needs to be set.

	<syntaxhighlight lang=bash>		<syntaxhighlight lang=bash>

AI on Demand: Token streaming

Sst-djo — Mon, 09 Nov 2026 15:29:30 GMT

Created page with "= Overview = This pages describes some useful commands on how to test token streaming. = Usage = We assume you've received your API key from us in our usual manner. == Usage - Inference with token-stream == Regular token streaming is difficult to read in the console. The expanded command processes the output into a far more digestible format on the console. Chat platforms like LibreChat include native mechanisms for handling token streaming or parsing the stream. <..."

New page

= Overview =
This pages describes some useful commands on how to test token streaming.

= Usage =
We assume you've received your API key from us in our usual manner.

== Usage - Inference with token-stream ==

Regular token streaming is difficult to read in the console.

The expanded command processes the output into a far more digestible format on the console.

Chat platforms like LibreChat include native mechanisms for handling token streaming or parsing the stream.

<syntaxhighlight lang=bash>
#Modell auswählen
MODEL='zai-org/GLM-5.3-Flash'

#Prompt eingeben
PROMPT='Hallo, bitte führe meine Prompt aus'

#Tokenlimit setzen
MAX_TOKENS=2000

#Inference Call
jq -n \
--arg model "$MODEL" \
--arg prompt "$PROMPT" \
--argjson max_tokens "$MAX_TOKENS" '{
model: $model,
messages: [
{
role: "user",
content: $prompt
}
],
stream: true,
stream_options: {include_usage: true},
max_tokens: $max_tokens
}' |
curl --fail-with-body --no-buffer -sS \
--connect-timeout 5 \
--max-time 180 \
https://llm.stoney-cloud.com/v1/chat/completions \
-H "Authorization: Bearer ${STONEY_KEY:?Set your production STONEY_KEY}" \
-H 'Content-Type: application/json' \
--data-binary @- |
tee "$STREAM_FILE" |
python3 -u -c '
import json
import sys

section = None
latest_usage = None
done = False

for line in sys.stdin:
if not line.startswith("data:"):
continue

payload = line[5:].strip()
if payload == "[DONE]":
done = True
continue

try:
event = json.loads(payload)
except json.JSONDecodeError:
continue

if not isinstance(event, dict):
continue

if "error" in event:
print("\nAPI error:", event["error"], flush=True)

for choice in event.get("choices") or []:
delta = choice.get("delta") or {}
reasoning = delta.get("reasoning") or delta.get("reasoning_content")

for label, text in (
("Reasoning", reasoning),
("Answer", delta.get("content")),
):
if isinstance(text, str) and text:
if section != label:
print("\n\n--- {} ---".format(label), flush=True)
section = label
print(text, end="", flush=True)

if choice.get("finish_reason"):
print(
"\n[Finish reason: {}]".format(choice["finish_reason"]),
flush=True,
)

usage = event.get("usage")
if isinstance(usage, dict):
latest_usage = usage

if latest_usage is not None:
label = "Usage" if done else "Last reported usage (stream incomplete)"
print("\n\n{}:\n{}".format(
label, json.dumps(latest_usage, indent=2)
), flush=True)

print(
"\n[Stream complete]" if done else "\n[Stream ended without DONE]",
flush=True,
)
'
</syntaxhighlight>

[[Category: AI on Demand]]

AI on Demand: Test streaming

Sst-djo — Fri, 06 Nov 2026 10:32:48 GMT

← Older revision		Revision as of 11:32, 6 November 2026
Line 11:		Line 11:
	* [[AI_on_demand:_apertus-ai/Apertus-v1.5-8B\|apertus-ai/Apertus-v1.5-8B]]		* [[AI_on_demand:_apertus-ai/Apertus-v1.5-8B\|apertus-ai/Apertus-v1.5-8B]]
	* [[AI_on_demand:_deepseek-ai/DeepSeek-V4-Flash-0731\|deepseek-ai/DeepSeek-V4-Flash-0731]]		* [[AI_on_demand:_deepseek-ai/DeepSeek-V4-Flash-0731\|deepseek-ai/DeepSeek-V4-Flash-0731]]
			* [[AI_on_demand:_google/gemma-4-31B-it\|google/gemma-4-31B-it]]
	* [[AI_on_demand:_MiniMaxAI/MiniMax-M2.5\|MiniMaxAI/MiniMax-M2.5]]		* [[AI_on_demand:_MiniMaxAI/MiniMax-M2.5\|MiniMaxAI/MiniMax-M2.5]]
	* [[AI_on_demand:_NVIDIA/NVIDIA-Nemotron-3-Super-120B-A12B\|NVIDIA/NVIDIA-Nemotron-3-Super-120B-A12B]]		* [[AI_on_demand:_NVIDIA/NVIDIA-Nemotron-3-Super-120B-A12B\|NVIDIA/NVIDIA-Nemotron-3-Super-120B-A12B]]

AI on demand: Configuring OpenCode Desktop

Sst-pde — Tue, 03 Nov 2026 15:06:29 GMT

Overview

← Older revision		Revision as of 16:06, 3 November 2026
Line 3:		Line 3:
	[https://opencode.ai OpenCode] is an open-source AI coding agent, available as a desktop application, in the browser or via the terminal. It supports tool calls, MCP servers, skills, plugins, etc. for software development workflows. OpenCode can be configured to exclusively use AI models hosted by [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG] via its OpenAI-compatible API.		[https://opencode.ai OpenCode] is an open-source AI coding agent, available as a desktop application, in the browser or via the terminal. It supports tool calls, MCP servers, skills, plugins, etc. for software development workflows. OpenCode can be configured to exclusively use AI models hosted by [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG] via its OpenAI-compatible API.

	'''Note:''' This guide walks through configuring the desktop variant of OpenCode. If you are looking for guidance on setting up the terminal version of OpenCode, see [[AI on ~~Demand~~: ~~Configure~~ OpenCode Terminal]]		'''Note:''' This guide walks through configuring the desktop variant of OpenCode. If you are looking for guidance on setting up the terminal version of OpenCode, see [[AI on demand: Configuring OpenCode Terminal]]

	The following steps describe how to create a basic OpenCode setup using models provided by stepping stone AG, and some good first steps to control and extend the agents capabilities.		The following steps describe how to create a basic OpenCode setup using models provided by stepping stone AG, and some good first steps to control and extend the agents capabilities.

AI on demand: Configuring OpenCode Terminal

Sst-pde — Tue, 03 Nov 2026 15:06:17 GMT

Overview

← Older revision		Revision as of 16:06, 3 November 2026
(One intermediate revision by the same user not shown)
Line 3:		Line 3:
	[https://opencode.ai OpenCode] is an open-source AI coding agent, available as a desktop application, in the browser or via the terminal. It supports tool calls, MCP servers, skills, plugins, etc. for software development workflows. OpenCode can be configured to exclusively use AI models hosted by [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG] via its OpenAI-compatible API.		[https://opencode.ai OpenCode] is an open-source AI coding agent, available as a desktop application, in the browser or via the terminal. It supports tool calls, MCP servers, skills, plugins, etc. for software development workflows. OpenCode can be configured to exclusively use AI models hosted by [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG] via its OpenAI-compatible API.

	'''Note:''' This guide walks through configuring the terminal variant of OpenCode. If you are looking for guidance on setting up the terminal version of OpenCode, see [[AI on ~~Demand~~: ~~Configure~~ OpenCode Desktop]]		'''Note:''' This guide walks through configuring the terminal variant of OpenCode. If you are looking for guidance on setting up the terminal version of OpenCode, see [[AI on demand: Configuring OpenCode Desktop]]

	The following steps describe how to create a basic OpenCode setup using models provided by stepping stone AG, and some good first steps to control and extend the agents capabilities.		The following steps describe how to create a basic OpenCode setup using models provided by stepping stone AG, and some good first steps to control and extend the agents capabilities.

Main Page

Sst-pde — Tue, 03 Nov 2026 15:05:47 GMT

Artificial Intelligence (AI)

← Older revision		Revision as of 16:05, 3 November 2026
Line 82:		Line 82:
	[[:Category: OpenCode]]		[[:Category: OpenCode]]
	* [[AI on demand: Configuring OpenCode Desktop]]		* [[AI on demand: Configuring OpenCode Desktop]]
			* [[AI on demand: Configuring OpenCode Terminal]]

	== Quick links ==		== Quick links ==

AI on demand: Configuring OpenCode Terminal

Sst-pde — Tue, 03 Nov 2026 15:05:33 GMT

Created page with "== Overview == [https://opencode.ai OpenCode] is an open-source AI coding agent, available as a desktop application, in the browser or via the terminal. It supports tool calls, MCP servers, skills, plugins, etc. for software development workflows. OpenCode can be configured to exclusively use AI models hosted by [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG] via its OpenAI-compatible API. '''Note:''' This guide walks through confi..."

New page

== Overview ==

[https://opencode.ai OpenCode] is an open-source AI coding agent, available as a desktop application, in the browser or via the terminal. It supports tool calls, MCP servers, skills, plugins, etc. for software development workflows. OpenCode can be configured to exclusively use AI models hosted by [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG] via its OpenAI-compatible API.

'''Note:''' This guide walks through configuring the terminal variant of OpenCode. If you are looking for guidance on setting up the terminal version of OpenCode, see [[AI on Demand: Configure OpenCode Desktop]]

The following steps describe how to create a basic OpenCode setup using models provided by stepping stone AG, and some good first steps to control and extend the agents capabilities.

{{Note|title=⚠️ Warning|content=Remember that agents such as OpenCode can autonomously perform actions on the system they run, such as reading / editing files, making web requests, executing terminal commands, etc. Agents may also interpret untrusted inputs such as skills or AGENTS.md files as instructions. Consider carefully which actions an agent should be allowed to execute, and whether a container or VM solution may be appropriate to prevent your agent from accessing sensitive information.}}

== Prerequisites ==

* A valid [https://www.stepping-stone.ch/en/products/artificial-intelligence AI on Demand] API key.
* Network access to <code>https://llm.stoney-cloud.com/v1</code>
* The ID of the AI models you wish to use with OpenCode (e.g.
* ''Optional:'' [https://nodejs.org Node.js] 20 or newer if you intend to install opencode via npm.

== Installation ==

Use your JavaScript package manager of choice to install the terminal variant of OpenCode:

<syntaxhighlight lang="bash">
npm install -g @opencode/cli
# or
bun install -g --trust @opencode/cli
# or
brew install anomalyco/tap/opencode-v2
</syntaxhighlight>

See [https://opencode.ai/v2/docs https://opencode.ai/v2/docs] for further information on how to install OpenCode.

Verify that the installation was successful by starting OpenCode:
<syntaxhighlight lang="bash">
opencode --version
opencode
</syntaxhighlight>

<gallery mode=packed heights=200>
OpenCode TUI.png|thumb|OpenCode Terminal
</gallery>

== Configure provider and models ==

To configure stepping stone AG as your provider, you must modify your OpenCode configuration file to include the following:

<syntaxhighlight lang="json">
{
"$schema": "https://opencode.ai/config.json",
"providers": {
"stepping-stone": {
"name": "stepping stone AG",
"package": "@opencode/ai/providers/openai-compatible",
"env": ["STONEY_KEY"],
"settings": {
"baseURL": "https://llm.stoney-cloud.com/v1"
},
"models": {}
}
}
}
</syntaxhighlight>

On Linux machines, the default location of your personal configuration file is <code>~/.config/opencode/opencode.jsonc</code> or equivalent. See [https://opencode.ai/v2/docs/config/#locations https://opencode.ai/v2/docs/config/#locations] for more information on where to place the configuration file on your system.

'''Note:''' When configuring stepping stone AG as your provider as above, OpenCode expects to find your API key in an environment variable called <code>STONEY_KEY</code>. Alternatively you may add your key directly to the configuration file by removing the <code>"env": ["STONEY_KEY"],</code> line, and adding your key as follows:

<syntaxhighlight lang="json">
...
"settings": {
"baseURL": "https://llm.stoney-cloud.com/v1",
"apiKey": "YOUR API KEY GOES HERE",
},
...
</syntaxhighlight>

{{Note|title=⚠️ Warning|content=Your API key must be kept secret by you! Be careful where you store your key to prevent it from being stolen or abused by a third party.}}

Next, you should configure at least one model you wish to use with OpenCode. The following is a configuration for the <code>GLM-5.3-Flash</code> model provided by stepping stone AG:

<syntaxhighlight lang="json">
...
"models": {
"zai-org/GLM-5.3-Flash": {
"name": "GLM 5.3 Flash",
"limit": {
"context": 262144,
},
"cost": {
"input": 0.3,
"output": 1.5,
"cache": {
"read": 0.045,
},
},
},
}
</syntaxhighlight>

Replace <code>zai-org/GLM-5.3-Flash</code> with the ID of the particular model you wish to use, and refer to [https://www.stepping-stone.ch/en/products/artificial-intelligence https://www.stepping-stone.ch/en/products/artificial-intelligence] for pricing and context window limits. Additional model specific configuration can be added as needed. See [https://opencode.ai/v2/docs/models/ https://opencode.ai/v2/docs/models/] for further information on configuring models.

To check if your configuration is correct, restart OpenCode after changing the configuration file. If the configuration is correct, you should be able to select the newly added model hosted by stepping stone AG from the list of available models. You can open a selection of all models available to you via the shortcut <code>CTRL + X</code>, <code>m</code>. You may need to scroll to the end of the displayed list.
<gallery mode=packed heights=200>
OpenCode_TUI_Model_selection.png|Model selection menu
</gallery>

Finally, select the newly added model, then enter a first prompt to test that the model is working correctly.
<gallery mode=packed heights=200>
OpenCode_TUI_Initial_prompt.png|Enter an initial prompt
OpenCode_TUI_Initial_prompt_response.png|Check that you receive a response
</gallery>

== Useful short-cuts and menus ==

The following is a list of useful shortcuts to get started with using OpenCode:

* <code>CTRL + P</code>: Open the main menu to interact with the agent, manage sessions and MCP servers, configuration, theming, etc.
* <code>CTRL + X</code> / <code>N</code>: Open a fresh session.
* <code>CTRL + X</code> / <code>L</code>: Switch to another existing session.
* <code>CTRL + X</code> / <code>W</code>: Close the current session tab.
* <code>CTRL + X</code> / <code>M</code>: Switch the current model.
* <code>CTRL + X</code> / <code>A</code>: Switch the current agent mode.
* <code>CTRL + X</code> / <code>Q</code>: Quit OpenCode.
* <code>CTRL + C</code>: Clear prompt input field.

== Permissions ==

OpenCode provides a built in sandbox to limit what actions an agent is allowed to perform. You can configure permissions via the configuration file (there is currently no option to configure permissions via the user interface). Permissions are defined as a list of rules. For each action (e.g. read/write a file, execute a shell command, etc.), you may define a set of rules, each with a pattern determining when the rule applies, and an effect to determine if the action is allowed, denied or requires explicit approval by the user. The following shows some examples of permission rules:

<syntaxhighlight lang="json">
...
"permissions": [
{ "action": "shell", "resource": "*", "effect": "ask" },
{ "action": "shell", "resource": "git status *", "effect": "allow" },
{ "action": "shell", "resource": "git push *", "effect": "deny" },
{ "action": "skill", "resource": "*", "effect": "allow" },
{ "action": "read", "resource": "*", "effect": "allow" },
{ "action": "edit", "resource": "*.html", "effect": "allow" },
]
...
</syntaxhighlight>

{{Note|title=⚠️ Warning|content=Try to keep the permissions you define as restrictive as feasable. In particular, do not allow your agent to perform sensitive or dangerous actions without your explicit approval.}}

{{Note|title=⚠️ Warning|content=While OpenCode provides a sandbox controlled via permissions, you should never rely exclusively on the sandbox to ensure that your agent does not perform dangerous actions or accesses sensitive data. Always exercise good judgement when deciding what your agent should be allowed to do.}}

== Further steps ==

You now have a basic OpenCode agent setup available to you. You may further customise and extend OpenCode in a variety of ways, such as by adding skills, configuring MCP servers or OpenCode plugins.

Check [https://opencode.ai/v2/docs/ https://opencode.ai/v2/docs/] to see how you can extend your configuration file to adapt your agent to your needs.

= References =

* [https://opencode.ai/v2/docs/ OpenCode documentation]
* [https://opencode.ai/v2/docs/providers OpenCode providers guide]
* [https://opencode.ai/v2/docs/config OpenCode configuration reference] and [https://opencode.ai/config.json configuration schema]
* [https://opencode.ai/v2/docs/permissions OpenCode permissions guide]
* [https://opencode.ai/v2/docs/mcp-servers OpenCode MCP servers guide]
* [https://modelcontextprotocol.io Model Context Protocol specification]
* [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG – Artificial Intelligence]
* [[AI on Demand: Usage]]

[[Category: AI on Demand]]
[[Category: OpenCode]]

File:OpenCode TUI Initial prompt response.png

Sst-pde — Tue, 03 Nov 2026 14:58:10 GMT

Sst-pde uploaded File:OpenCode TUI Initial prompt response.png

New page

File:OpenCode TUI Initial prompt.png

Sst-pde — Tue, 03 Nov 2026 14:56:54 GMT

Sst-pde uploaded File:OpenCode TUI Initial prompt.png

New page

File:OpenCode TUI Model selection.png

Sst-pde — Tue, 03 Nov 2026 14:54:53 GMT

Sst-pde uploaded File:OpenCode TUI Model selection.png

New page

File:OpenCode TUI.png

Sst-pde — Tue, 03 Nov 2026 14:52:00 GMT

Sst-pde uploaded a new version of File:OpenCode TUI.png

New page

OpenCode TUI adter startup

Main Page

Sst-pde — Tue, 03 Nov 2026 14:46:38 GMT

Artificial Intelligence (AI)

← Older revision		Revision as of 15:46, 3 November 2026
Line 79:		Line 79:
	* [[AI on demand: jaredpalmer/kev-27b]]		* [[AI on demand: jaredpalmer/kev-27b]]
	* [[AI on demand: Mapika/decider-31b]]		* [[AI on demand: Mapika/decider-31b]]

			[[:Category: OpenCode]]
			* [[AI on demand: Configuring OpenCode Desktop]]

	== Quick links ==		== Quick links ==

AI on Demand: Configuring OpenCode Desktop

Sst-pde — Tue, 03 Nov 2026 14:46:12 GMT

Sst-pde moved page AI on Demand: Configuring OpenCode Desktop to AI on demand: Configuring OpenCode Desktop

← Older revision		Revision as of 15:46, 3 November 2026
(One intermediate revision by the same user not shown)
Line 148:		Line 148:

	[[Category: AI on Demand]]		[[Category: AI on Demand]]
			[[Category: OpenCode]]

Category:OpenCode

Sst-pde — Tue, 03 Nov 2026 14:44:39 GMT

Created page with "Category: AI on Demand"

New page

[[Category: AI on Demand]]

AI on Demand: Configuring OpenCode

Sst-pde — Tue, 03 Nov 2026 14:43:52 GMT

Sst-pde moved page AI on Demand: Configuring OpenCode to AI on Demand: Configuring OpenCode Desktop

New page

== Overview ==

[https://opencode.ai OpenCode] is an open-source AI coding agent, available as a desktop application, in the browser or via the terminal. It supports tool calls, MCP servers, skills, plugins, etc. for software development workflows. OpenCode can be configured to exclusively use AI models hosted by [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG] via its OpenAI-compatible API.

'''Note:''' This guide walks through configuring the desktop variant of OpenCode. If you are looking for guidance on setting up the terminal version of OpenCode, see [[AI on Demand: Configure OpenCode Terminal]]

The following steps describe how to create a basic OpenCode setup using models provided by stepping stone AG, and some good first steps to control and extend the agents capabilities.

{{Note|title=⚠️ Warning|content=Remember that agents such as OpenCode can autonomously perform actions on the system they run, such as reading / editing files, making web requests, executing terminal commands, etc. Agents may also interpret untrusted inputs such as skills or AGENTS.md files as instructions. Consider carefully which actions an agent should be allowed to execute, and whether a container or VM solution may be appropriate to prevent your agent from accessing sensitive information.}}

== Prerequisites ==

* A valid [https://www.stepping-stone.ch/en/products/artificial-intelligence AI on Demand] API key.
* Network access to <code>https://llm.stoney-cloud.com/v1</code>
* The ID of the AI models you wish to use with OpenCode (e.g.
* ''Optional:'' [https://nodejs.org Node.js] 20 or newer if you intend to install opencode via npm.

== Installation ==

Download and install the [https://opencode.ai/v2/docs#desktop binary / package] appropriate for your system.

Once complete, open the OpenCode desktop app to verify that the installation was successful

<gallery mode=packed heights=200px>
OpenCode Desktop homepage.png|OpenCode Desktop
</gallery>

== Configure provider and models ==

OpenCode desktop currently does not support adding custom model providers directly from its user interface. You will need to configure stepping stone AG as your provider by modifying your OpenCode configuration file to include the following:

<syntaxhighlight lang="json">
{
"$schema": "https://opencode.ai/config.json",
"providers": {
"stepping-stone": {
"name": "stepping stone AG",
"package": "@opencode/ai/providers/openai-compatible",
"env": ["STONEY_KEY"],
"settings": {
"baseURL": "https://llm.stoney-cloud.com/v1"
},
"models": {}
}
}
}
</syntaxhighlight>

On Linux machines, the default location of your personal configuration file is <code>~/.config/opencode/opencode.jsonc</code> or equivalent. See [https://opencode.ai/v2/docs/config/#locations https://opencode.ai/v2/docs/config/#locations] for more information on where to place the configuration file on your system.

'''Note:''' When configuring stepping stone AG as your provider as above, OpenCode expects to find your API key in an environment variable called <code>STONEY_KEY</code>. Alternatively you may add your key directly to the configuration file by removing the <code>"env": ["STONEY_KEY"],</code> line, and adding your key as follows:

<syntaxhighlight lang="json">
...
"settings": {
"baseURL": "https://llm.stoney-cloud.com/v1",
"apiKey": "YOUR API KEY GOES HERE",
},
...
</syntaxhighlight>

{{Note|title=⚠️ Warning|content=Your API key must be kept secret by you! Be careful where you store your key to prevent it from being stolen or abused by a third party.}}

Next, you should configure at least one model you wish to use with OpenCode. The following is a configuration for the <code>GLM-5.3-Flash</code> model provided by stepping stone AG:

<syntaxhighlight lang="json">
...
"models": {
"zai-org/GLM-5.3-Flash": {
"name": "GLM 5.3 Flash",
"limit": {
"context": 262144,
},
"cost": {
"input": 0.3,
"output": 1.5,
"cache": {
"read": 0.045,
},
},
},
}
</syntaxhighlight>

Replace <code>zai-org/GLM-5.3-Flash</code> with the ID of the particular model you wish to use, and refer to [https://www.stepping-stone.ch/en/products/artificial-intelligence https://www.stepping-stone.ch/en/products/artificial-intelligence] for pricing and context window limits. Additional model specific configuration can be added as needed. See [https://opencode.ai/v2/docs/models/ https://opencode.ai/v2/docs/models/] for further information on configuring models.

To check if your configuration is correct, restart OpenCode after changing the configuration file. If the configuration is correct, stepping stone AG should appear as a configured provider, and your selected model in the available models list.

<gallery mode=packed heights=200>
OpenCode_Home_button.png|1. Open home page
OpenCode_settings_menu.png|2. Open settings menu
OpenCode_Desktop_Stepping_Stone_Provider.png|3. Check that stepping stone AG appears as a provider
OpenCode_Desktop_Stepping_Stone_Model.png|4. Check that your selected model appears at the bottom of the list
</gallery>

Finally, start a new session, select the model you have just configured and enter a first prompt to test that the model is working correctly.

<gallery mode=packed heights=200>
OpenCode_Desktop_Stepping_Stone_Model_selected.png|1. Start a new session and select a stepping stone AG provided model
OpenCode_Desktop_First_Prompt.png|2. Enter a first prompt to check that the configuration works
</gallery>

== Configure appearance and notifications ==

You may customise the appearance and layout of the UI, notification settings, shortcuts and more via the settings menu.

<gallery mode=packed heights=200>
OpenCode_Desktop_settings.png|Available settings submenus
</gallery>

== Permissions ==

OpenCode provides a built in sandbox to limit what actions an agent is allowed to perform. You can configure permissions via the configuration file (there is currently no option to configure permissions via the user interface). Permissions are defined as a list of rules. For each action (e.g. read/write a file, execute a shell command, etc.), you may define a set of rules, each with a pattern determining when the rule applies, and an effect to determine if the action is allowed, denied or requires explicit approval by the user. The following shows some examples of permission rules:

<syntaxhighlight lang="json">
...
"permissions": [
{ "action": "shell", "resource": "*", "effect": "ask" },
{ "action": "shell", "resource": "git status *", "effect": "allow" },
{ "action": "shell", "resource": "git push *", "effect": "deny" },
{ "action": "skill", "resource": "*", "effect": "allow" },
{ "action": "read", "resource": "*", "effect": "allow" },
{ "action": "edit", "resource": "*.html", "effect": "allow" },
]
...
</syntaxhighlight>

{{Note|title=⚠️ Warning|content=Try to keep the permissions you define as restrictive as feasable. In particular, do not allow your agent to perform sensitive or dangerous actions without your explicit approval.}}

{{Note|title=⚠️ Warning|content=While OpenCode provides a sandbox controlled via permissions, you should never rely exclusively on the sandbox to ensure that your agent does not perform dangerous actions or accesses sensitive data. Always exercise good judgement when deciding what your agent should be allowed to do.}}

== Further steps ==

You now have a basic OpenCode agent setup available to you. You may further customise and extend OpenCode in a variety of ways, such as by adding skills, configuring MCP servers or OpenCode plugins.

Check [https://opencode.ai/v2/docs/ https://opencode.ai/v2/docs/] or the settings menu of OpenCode to see how you can extend your agent to your needs.

= References =

* [https://opencode.ai/v2/docs/ OpenCode documentation]
* [https://opencode.ai/v2/docs/providers OpenCode providers guide]
* [https://opencode.ai/v2/docs/config OpenCode configuration reference] and [https://opencode.ai/config.json configuration schema]
* [https://opencode.ai/v2/docs/permissions OpenCode permissions guide]
* [https://opencode.ai/v2/docs/mcp-servers OpenCode MCP servers guide]
* [https://modelcontextprotocol.io Model Context Protocol specification]
* [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG – Artificial Intelligence]
* [[AI on Demand: Usage]]

[[Category: AI on Demand]]

AI on Demand: Configuring OpenCode

Sst-pde — Tue, 03 Nov 2026 14:43:44 GMT

New page

== Overview ==

[https://opencode.ai OpenCode] is an open-source AI coding agent, available as a desktop application, in the browser or via the terminal. It supports tool calls, MCP servers, skills, plugins, etc. for software development workflows. OpenCode can be configured to exclusively use AI models hosted by [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG] via its OpenAI-compatible API.

'''Note:''' This guide walks through configuring the desktop variant of OpenCode. If you are looking for guidance on setting up the terminal version of OpenCode, see [[AI on Demand: Configure OpenCode Terminal]]

The following steps describe how to create a basic OpenCode setup using models provided by stepping stone AG, and some good first steps to control and extend the agents capabilities.

{{Note|title=⚠️ Warning|content=Remember that agents such as OpenCode can autonomously perform actions on the system they run, such as reading / editing files, making web requests, executing terminal commands, etc. Agents may also interpret untrusted inputs such as skills or AGENTS.md files as instructions. Consider carefully which actions an agent should be allowed to execute, and whether a container or VM solution may be appropriate to prevent your agent from accessing sensitive information.}}

== Prerequisites ==

* A valid [https://www.stepping-stone.ch/en/products/artificial-intelligence AI on Demand] API key.
* Network access to <code>https://llm.stoney-cloud.com/v1</code>
* The ID of the AI models you wish to use with OpenCode (e.g.
* ''Optional:'' [https://nodejs.org Node.js] 20 or newer if you intend to install opencode via npm.

== Installation ==

Download and install the [https://opencode.ai/v2/docs#desktop binary / package] appropriate for your system.

Once complete, open the OpenCode desktop app to verify that the installation was successful

<gallery mode=packed heights=200px>
OpenCode Desktop homepage.png|OpenCode Desktop
</gallery>

== Configure provider and models ==

OpenCode desktop currently does not support adding custom model providers directly from its user interface. You will need to configure stepping stone AG as your provider by modifying your OpenCode configuration file to include the following:

<syntaxhighlight lang="json">
{
"$schema": "https://opencode.ai/config.json",
"providers": {
"stepping-stone": {
"name": "stepping stone AG",
"package": "@opencode/ai/providers/openai-compatible",
"env": ["STONEY_KEY"],
"settings": {
"baseURL": "https://llm.stoney-cloud.com/v1"
},
"models": {}
}
}
}
</syntaxhighlight>

On Linux machines, the default location of your personal configuration file is <code>~/.config/opencode/opencode.jsonc</code> or equivalent. See [https://opencode.ai/v2/docs/config/#locations https://opencode.ai/v2/docs/config/#locations] for more information on where to place the configuration file on your system.

'''Note:''' When configuring stepping stone AG as your provider as above, OpenCode expects to find your API key in an environment variable called <code>STONEY_KEY</code>. Alternatively you may add your key directly to the configuration file by removing the <code>"env": ["STONEY_KEY"],</code> line, and adding your key as follows:

<syntaxhighlight lang="json">
...
"settings": {
"baseURL": "https://llm.stoney-cloud.com/v1",
"apiKey": "YOUR API KEY GOES HERE",
},
...
</syntaxhighlight>

{{Note|title=⚠️ Warning|content=Your API key must be kept secret by you! Be careful where you store your key to prevent it from being stolen or abused by a third party.}}

Next, you should configure at least one model you wish to use with OpenCode. The following is a configuration for the <code>GLM-5.3-Flash</code> model provided by stepping stone AG:

<syntaxhighlight lang="json">
...
"models": {
"zai-org/GLM-5.3-Flash": {
"name": "GLM 5.3 Flash",
"limit": {
"context": 262144,
},
"cost": {
"input": 0.3,
"output": 1.5,
"cache": {
"read": 0.045,
},
},
},
}
</syntaxhighlight>

Replace <code>zai-org/GLM-5.3-Flash</code> with the ID of the particular model you wish to use, and refer to [https://www.stepping-stone.ch/en/products/artificial-intelligence https://www.stepping-stone.ch/en/products/artificial-intelligence] for pricing and context window limits. Additional model specific configuration can be added as needed. See [https://opencode.ai/v2/docs/models/ https://opencode.ai/v2/docs/models/] for further information on configuring models.

To check if your configuration is correct, restart OpenCode after changing the configuration file. If the configuration is correct, stepping stone AG should appear as a configured provider, and your selected model in the available models list.

<gallery mode=packed heights=200>
OpenCode_Home_button.png|1. Open home page
OpenCode_settings_menu.png|2. Open settings menu
OpenCode_Desktop_Stepping_Stone_Provider.png|3. Check that stepping stone AG appears as a provider
OpenCode_Desktop_Stepping_Stone_Model.png|4. Check that your selected model appears at the bottom of the list
</gallery>

Finally, start a new session, select the model you have just configured and enter a first prompt to test that the model is working correctly.

<gallery mode=packed heights=200>
OpenCode_Desktop_Stepping_Stone_Model_selected.png|1. Start a new session and select a stepping stone AG provided model
OpenCode_Desktop_First_Prompt.png|2. Enter a first prompt to check that the configuration works
</gallery>

== Configure appearance and notifications ==

You may customise the appearance and layout of the UI, notification settings, shortcuts and more via the settings menu.

<gallery mode=packed heights=200>
OpenCode_Desktop_settings.png|Available settings submenus
</gallery>

== Permissions ==

OpenCode provides a built in sandbox to limit what actions an agent is allowed to perform. You can configure permissions via the configuration file (there is currently no option to configure permissions via the user interface). Permissions are defined as a list of rules. For each action (e.g. read/write a file, execute a shell command, etc.), you may define a set of rules, each with a pattern determining when the rule applies, and an effect to determine if the action is allowed, denied or requires explicit approval by the user. The following shows some examples of permission rules:

<syntaxhighlight lang="json">
...
"permissions": [
{ "action": "shell", "resource": "*", "effect": "ask" },
{ "action": "shell", "resource": "git status *", "effect": "allow" },
{ "action": "shell", "resource": "git push *", "effect": "deny" },
{ "action": "skill", "resource": "*", "effect": "allow" },
{ "action": "read", "resource": "*", "effect": "allow" },
{ "action": "edit", "resource": "*.html", "effect": "allow" },
]
...
</syntaxhighlight>

{{Note|title=⚠️ Warning|content=Try to keep the permissions you define as restrictive as feasable. In particular, do not allow your agent to perform sensitive or dangerous actions without your explicit approval.}}

{{Note|title=⚠️ Warning|content=While OpenCode provides a sandbox controlled via permissions, you should never rely exclusively on the sandbox to ensure that your agent does not perform dangerous actions or accesses sensitive data. Always exercise good judgement when deciding what your agent should be allowed to do.}}

== Further steps ==

You now have a basic OpenCode agent setup available to you. You may further customise and extend OpenCode in a variety of ways, such as by adding skills, configuring MCP servers or OpenCode plugins.

Check [https://opencode.ai/v2/docs/ https://opencode.ai/v2/docs/] or the settings menu of OpenCode to see how you can extend your agent to your needs.

= References =

* [https://opencode.ai/v2/docs/ OpenCode documentation]
* [https://opencode.ai/v2/docs/providers OpenCode providers guide]
* [https://opencode.ai/v2/docs/config OpenCode configuration reference] and [https://opencode.ai/config.json configuration schema]
* [https://opencode.ai/v2/docs/permissions OpenCode permissions guide]
* [https://opencode.ai/v2/docs/mcp-servers OpenCode MCP servers guide]
* [https://modelcontextprotocol.io Model Context Protocol specification]
* [https://www.stepping-stone.ch/en/products/artificial-intelligence stepping stone AG – Artificial Intelligence]
* [[AI on Demand: Usage]]

[[Category: AI on Demand]]

File:OpenCode Desktop settings.png

Sst-pde — Tue, 03 Nov 2026 14:22:26 GMT

Sst-pde uploaded File:OpenCode Desktop settings.png

New page

File:OpenCode Desktop First Prompt.png

Sst-pde — Tue, 03 Nov 2026 14:19:37 GMT

Sst-pde uploaded File:OpenCode Desktop First Prompt.png

New page

File:OpenCode Desktop Stepping Stone Model selected.png

Sst-pde — Tue, 03 Nov 2026 14:17:52 GMT

Sst-pde uploaded File:OpenCode Desktop Stepping Stone Model selected.png

New page

File:OpenCode Desktop Stepping Stone Model.png

Sst-pde — Tue, 03 Nov 2026 14:11:09 GMT

Sst-pde uploaded File:OpenCode Desktop Stepping Stone Model.png

New page

File:OpenCode Desktop Stepping Stone Provider.png

Sst-pde — Tue, 03 Nov 2026 14:10:16 GMT

Sst-pde uploaded File:OpenCode Desktop Stepping Stone Provider.png

New page

File:OpenCode settings menu.png

Sst-pde — Tue, 03 Nov 2026 12:04:17 GMT

Sst-pde uploaded a new version of File:OpenCode settings menu.png

New page

OpenCode settings menu location highlighted

File:OpenCode Home button.png

Sst-pde — Tue, 03 Nov 2026 12:03:03 GMT

Sst-pde uploaded a new version of File:OpenCode Home button.png

New page

OpenCode Home button location highlighted

File:OpenCode custom provider.png

Sst-pde — Tue, 03 Nov 2026 10:42:29 GMT

Sst-pde uploaded File:OpenCode custom provider.png

New page

OpenCode custom provider location highlighted

File:OpenCode settings menu.png

Sst-pde — Tue, 03 Nov 2026 10:41:42 GMT

Sst-pde uploaded File:OpenCode settings menu.png

New page

OpenCode settings menu location highlighted

File:OpenCode Home button.png

Sst-pde — Tue, 03 Nov 2026 10:34:56 GMT

Sst-pde uploaded File:OpenCode Home button.png

New page

OpenCode Home button location highlighted

File:OpenCode TUI.png

Sst-pde — Tue, 03 Nov 2026 10:22:39 GMT

Sst-pde uploaded File:OpenCode TUI.png

New page

OpenCode TUI adter startup

File:OpenCode Desktop homepage.png

Sst-pde — Tue, 03 Nov 2026 10:13:29 GMT

Sst-pde uploaded File:OpenCode Desktop homepage.png

New page

OpenCode Desktop window after startup

AI on Demand: Configuring OpenCode

Sst-pde — Tue, 03 Nov 2026 09:51:35 GMT

Sst-pde deleted page AI on Demand: Configuring OpenCode Not ready for publishing yet

AI on demand: lightonai/LightOnOCR-2-1B

Sst-mei — Sat, 31 Oct 2026 04:58:25 GMT

Calling the model

← Older revision		Revision as of 05:58, 31 October 2026
Line 4:		Line 4:
	# Set your personal key:		# Set your personal key:
	STONEY_KEY=sk-...		STONEY_KEY=sk-...

			# Set the desired document to run OCR against:
			FILE_PATH=document.png

	# Set the desired model:		# Set the desired model:
	MODEL=lightonai/LightOnOCR-2-1B		MODEL=lightonai/LightOnOCR-2-1B

	~~# Set the desired document to run OCR against:~~
	~~FILE_PATH=document.png~~

	# Encode document contents with Base64 and detect MIME type:		# Encode document contents with Base64 and detect MIME type:

AI on demand: allenai/olmOCR-2-7B

Sst-mei — Sat, 31 Oct 2026 04:57:59 GMT

Calling the model

← Older revision		Revision as of 05:57, 31 October 2026
Line 4:		Line 4:
	# Set your personal key:		# Set your personal key:
	STONEY_KEY=sk-...		STONEY_KEY=sk-...

			# Set the desired document to run OCR against:
			FILE_PATH=document.png

	# Set the desired model:		# Set the desired model:
	MODEL=allenai/olmOCR-2-7B		MODEL=allenai/olmOCR-2-7B

	~~# Set the desired document to run OCR against:~~
	~~FILE_PATH=document.png~~

	# Encode document contents with Base64 and detect MIME type:		# Encode document contents with Base64 and detect MIME type:

AI on demand: opendatalab/MinerU2.5-2509-1.2B

Sst-mei — Sat, 31 Oct 2026 04:56:40 GMT

Calling the model

← Older revision		Revision as of 05:56, 31 October 2026
(2 intermediate revisions by the same user not shown)
Line 3:		Line 3:
	# Set your personal key:		# Set your personal key:
	STONEY_KEY=sk-...		STONEY_KEY=sk-...

			# Set the desired document to run OCR against:
			FILE_PATH=document.png

	# Set the desired model:		# Set the desired model:
	MODEL=opendatalab/MinerU2.5-2509-1.2B		MODEL=opendatalab/MinerU2.5-2509-1.2B

	~~# Set the desired document to run OCR against:~~
	~~FILE_PATH=document.png~~

	# Encode document contents with Base64 and detect MIME type:		# Encode document contents with Base64 and detect MIME type:

Security group rules

Sst-aba — Fri, 30 Oct 2026 11:47:36 GMT

← Older revision		Revision as of 12:47, 30 October 2026
(3 intermediate revisions by the same user not shown)
Line 152:		Line 152:

	\|-\|OpenStack Dashboard=		\|-\|OpenStack Dashboard=
	~~Work in progress!!~~
	= Managed security groups using the Dashboard =		= Managed security groups using the Dashboard =
	This chapter shows guides you through the daily operations regarding security groups in stepping stone AG's OpenStack based cloud.		This chapter shows guides you through the daily operations regarding security groups in stepping stone AG's OpenStack based cloud.
Line 180:		Line 179:
	In the left menu, choose <code>Project</code> → <code>Network</code> → <code>Security Groups</code>.		In the left menu, choose <code>Project</code> → <code>Network</code> → <code>Security Groups</code>.

	Next to the security group you want to add a rule to, click <code>Manage Rules</code> on the right~~. On the page that opens, click <code>+ Add Rule</code> and fill out the form with the values for the new rule~~.		Next to the security group you want to add a rule to, click <code>Manage Rules</code> on the right.

			[[File:Security group rules Managed security groups using the Dashboard 05 1440x900.png\|1000px]]


			On the page that opens, click <code>+ Add Rule</code> and fill out the form with the values for the new rule.

	~~[[File:Security group rules Managed security groups using the Dashboard 05 1440x900.png\|1000px]]~~

	[[File:Security group rules Managed security groups using the Dashboard 06 1440x900.png\|1000px]]		[[File:Security group rules Managed security groups using the Dashboard 06 1440x900.png\|1000px]]
Line 204:		Line 206:
	*** '''Ether Type''': Either <code>IPv4</code> or <code>IPv6</code>.		*** '''Ether Type''': Either <code>IPv4</code> or <code>IPv6</code>.

	== ~~Managed~~ security groups using the Dashboard - Assign to server ==		== Managing security groups using the Dashboard - Assign to server ==
			To assign a security group to an instance:

			In the left menu, choose <code>Project</code> → <code>Compute</code> → <code>Instances</code>.

			Next to the instance you want to assign the security group to, click the drop-down menu in the <code>Actions</code> column on the right and select <code>Edit Security Groups</code>.


			[[File:Security group rules Managed security groups using the Dashboard 09 1440x900.png\|1000px]]

			A pop-up window opens. On the left, it shows a list of <code>All Security Groups</code>; on the right, it shows the <code>Instance Security Groups</code>.

			To assign a security group to the instance, click <code>+</code> next to it in the list on the left. To remove a security group from the instance, click <code>-</code> next to it in the list on the right.


			[[File:Security group rules Managed security groups using the Dashboard 10 1440x900.png\|1000px]]
	== Managed security groups using the Dashboard - List ==		== Managed security groups using the Dashboard - List ==
	In the OpenStack Dashboard, you can list the security groups of a project. However, you cannot list the security groups of all projects at once. To do this, use the CLI as described in <code>Managing security groups using the CLI</code>.		In the OpenStack Dashboard, you can list the security groups of a project. However, you cannot list the security groups of all projects at once. To do this, use the CLI as described in <code>Managing security groups using the CLI</code>.

File:Security group rules Managed security groups using the Dashboard 10 1440x900.png

Sst-aba — Fri, 30 Oct 2026 11:03:42 GMT

Sst-aba uploaded File:Security group rules Managed security groups using the Dashboard 10 1440x900.png

New page

This screenshot shows where to click in order to assign or remove a security group to or from a server.

File:Security group rules Managed security groups using the Dashboard 09 1440x900.png

Sst-aba — Fri, 30 Oct 2026 11:02:17 GMT

Sst-aba uploaded File:Security group rules Managed security groups using the Dashboard 09 1440x900.png

New page

This screenshot shows where to click in order to get to the menu in which we can assign or remove security groups to or from a server.

Security group rules

Sst-aba — Fri, 30 Oct 2026 10:50:22 GMT

← Older revision		Revision as of 11:50, 30 October 2026
(One intermediate revision by the same user not shown)
Line 206:		Line 206:
	== Managed security groups using the Dashboard - Assign to server ==		== Managed security groups using the Dashboard - Assign to server ==
	== Managed security groups using the Dashboard - List ==		== Managed security groups using the Dashboard - List ==
			In the OpenStack Dashboard, you can list the security groups of a project. However, you cannot list the security groups of all projects at once. To do this, use the CLI as described in <code>Managing security groups using the CLI</code>.

			To list the security groups of a project:

			In the list at the top, select the project.

			In the left menu, choose <code>Project</code> → <code>Network</code> → <code>Security Groups</code>.


			[[File:Security group rules Managed security groups using the Dashboard 03 1440x900.png\|1000px]]

	== Managing security groups using the Dashboard - Show ==		== Managing security groups using the Dashboard - Show ==
	To view the properties of a security group:		To view the properties of a security group:

File:Security group rules Managed security groups using the Dashboard 03 1440x900.png

Sst-aba — Fri, 30 Oct 2026 10:49:48 GMT

Sst-aba uploaded File:Security group rules Managed security groups using the Dashboard 03 1440x900.png

New page

This screenshot shows how to get the list of security groups of a project using the OpenStack Dashboard.

Security group rules

Sst-aba — Fri, 30 Oct 2026 10:43:05 GMT

← Older revision		Revision as of 11:43, 30 October 2026
(2 intermediate revisions by the same user not shown)
Line 158:		Line 158:

	'''Note:''' This guide assumes that you already have access to the [https://dashboard.stoney-cloud.com OpenStack Dashboard] and are authenticated. If you are not, follow the [[Quickstart\|OpenStack quick start guide]].		'''Note:''' This guide assumes that you already have access to the [https://dashboard.stoney-cloud.com OpenStack Dashboard] and are authenticated. If you are not, follow the [[Quickstart\|OpenStack quick start guide]].
	== Managing security groups using the Dashboard – Create a group ==		== Managing security groups using the Dashboard - Create group ==
	A new security group can be created as follows:		A new security group can be created as follows:

Line 181:		Line 181:

	Next to the security group you want to add a rule to, click <code>Manage Rules</code> on the right. On the page that opens, click <code>+ Add Rule</code> and fill out the form with the values for the new rule.		Next to the security group you want to add a rule to, click <code>Manage Rules</code> on the right. On the page that opens, click <code>+ Add Rule</code> and fill out the form with the values for the new rule.


	[[File:Security group rules Managed security groups using the Dashboard 05 1440x900.png\|1000px]]		[[File:Security group rules Managed security groups using the Dashboard 05 1440x900.png\|1000px]]
Line 224:		Line 225:

	Alternatively, select the checkboxes on the left of the security groups you want to delete, then click the red <code>Delete Security Groups</code> button.		Alternatively, select the checkboxes on the left of the security groups you want to delete, then click the red <code>Delete Security Groups</code> button.


	[[File:Security group rules Managed security groups using the Dashboard 04 1440x900.png\|1000px]]		[[File:Security group rules Managed security groups using the Dashboard 04 1440x900.png\|1000px]]

← Older revision		Revision as of 11:32, 6 November 2026
Line 11:		Line 11:
	* [[AI_on_demand:_apertus-ai/Apertus-v1.5-8B\|apertus-ai/Apertus-v1.5-8B]]		* [[AI_on_demand:_apertus-ai/Apertus-v1.5-8B\|apertus-ai/Apertus-v1.5-8B]]
	* [[AI_on_demand:_deepseek-ai/DeepSeek-V4-Flash-0731\|deepseek-ai/DeepSeek-V4-Flash-0731]]		* [[AI_on_demand:_deepseek-ai/DeepSeek-V4-Flash-0731\|deepseek-ai/DeepSeek-V4-Flash-0731]]
			* [[AI_on_demand:_google/gemma-4-31B-it\|google/gemma-4-31B-it]]
	* [[AI_on_demand:_MiniMaxAI/MiniMax-M2.5\|MiniMaxAI/MiniMax-M2.5]]		* [[AI_on_demand:_MiniMaxAI/MiniMax-M2.5\|MiniMaxAI/MiniMax-M2.5]]
	* [[AI_on_demand:_NVIDIA/NVIDIA-Nemotron-3-Super-120B-A12B\|NVIDIA/NVIDIA-Nemotron-3-Super-120B-A12B]]		* [[AI_on_demand:_NVIDIA/NVIDIA-Nemotron-3-Super-120B-A12B\|NVIDIA/NVIDIA-Nemotron-3-Super-120B-A12B]]