For developers
An OpenAI-compatible API on your own hardware
Point the OpenAI SDK at your own computers, or run Claude Code, Codex, or OpenCode as a local coding agent on a model there. No per-token bill, no rate limits but your hardware, and source code, customer data, and prompts that never leave the building.
Building an app on Toskar itself, with its chats, knowledge, and automations? Use the official SDKs.
What you get
Built to be called from code
An OpenAI-compatible API
Change the base URL and keep your SDK. Chat completions with streaming, JSON and JSON-schema output, and reasoning effort, answered by models on your hardware.
Coding agents on local models
Claude Code and other agents built for Anthropic's Messages API, and Codex-style tools that use OpenAI's Responses API, run on a model on your computer, with their own tools passed through. toskarctl connect sets up Claude Code, Codex, OpenCode, and aider.
Your models in your editor
Claude Desktop, Claude Code, Cursor, and VS Code can ask your local models and search your connected docs over MCP. API Access gives the settings to paste.
A key per developer
Each key has its own limits on memory, knowledge, and tools, and can be rotated or revoked. Nothing needs a key on the computer itself.
Coding models and tools
Qwen 2.5 Coder and other coding models from the catalog, and a Programming profile that can read files, run commands, and use git when you allow it.
Several machines, one endpoint
Pair a GPU box with your laptop. Toskar places each request on the computer that can run it, behind the same address.
Open source, and yours to change
Toskar Core is AGPL-3.0 Go and TypeScript, with an OpenAPI spec. Read it, run it in CI, or build on it with the official SDKs.
Try it
Call it like you call OpenAI
Your local AI inside the apps you use
These apps can ask your local AI and search your connected knowledge over MCP. Toskar gives you the settings to paste.
- Claude Desktop
- Claude Code
- Cursor
- VS Code
- any MCP app
Anything that speaks the OpenAI API
Point any app or SDK that lets you set an OpenAI base URL at Toskar. Use auto as the model and Toskar picks one for each request. No key is needed on this computer.
from openai import OpenAI
client = OpenAI(base_url="http://127.0.0.1:7331/v1", api_key="local")
reply = client.chat.completions.create(
model="auto",
messages=[{"role": "user", "content": "Hello!"}],
)
print(reply.choices[0].message.content)With Toskar running, this prints a reply from a model on your computer.
For a team
One box for the whole team
Step 1
Put Toskar on a machine with memory
A Mac with 64 GB of unified memory, or a PC with 32 GB of memory and a 24 GB graphics card, runs every model in the catalog, coding models included.
Step 2
Open it to the network
Turn on local network access and give each developer a key, limited to what their tools need.
Step 3
Use it from everything
SDKs and scripts over the OpenAI API, editors over MCP, and the Toskar chat itself on that machine.
Good to know
Where it differs from OpenAI's API
- Your tools go to raw: models. On /v1/chat/completions, tool definitions reach the model only when you name it as
raw:and its ID; otherwise Toskar's own harness answers, with the tools set in its profile. What's an agent harness? - Chat, messages, and responses. There is no embeddings, images, or audio endpoint.
- Some parameters are ignored.
temperatureandmax_tokensare accepted and not applied, and a model must be installed before a request can use it. - HTTP on the local network. Put a TLS proxy in front of it before reaching it from anywhere you don't trust.
The full list is in the API docs, and the code is on GitHub.
From the blog
Guides for developers
- Use Toskar from your phone, laptop, and other appsRun Toskar on the computer with the most memory, then use it from every device on your home or office network. Four steps, about five minutes.October 5, 2026 · 4 min read
- What can my computer run? A plain guide to local AI and memoryMemory decides which AI models your computer can run. How much each model size needs, what fits in 8, 16, 32, and 64 GB, and Mac vs. PC.October 5, 2026 · 4 min read
- Automations: let Toskar check things for youHave your own AI watch prices, sum up the news, and track releases on a schedule, telling you only when it matters. A tutorial with screenshots.October 5, 2026 · 6 min read
What to get
The products that fit
Questions
- Which OpenAI endpoints does Toskar support?
- GET /v1/models and POST /v1/chat/completions, with streaming, response_format (json_object and json_schema), and reasoning_effort, plus POST /v1/responses for agents that use OpenAI's Responses format. Anthropic's POST /v1/messages and /v1/messages/count_tokens are there too. There is no embeddings endpoint yet; document search runs inside Toskar.
- Can I pass my own tools (function calling)?
- Yes. Send "model": "raw:<model id>" to /v1/chat/completions and the request goes to the model as it came, with your tools, tool_choice, stop, and logprobs, and the model's tool_calls come back, streamed or not. /v1/messages and /v1/responses always work this way. From the computer itself no key is needed; for a key, an Admin turns on Drive models directly, for coding agents, in its permissions. Without raw, Toskar's own assistant answers with the tools set in its profile.
- Can Cursor or Claude Code use Toskar as their main model?
- Claude Code can: point its base URL at Toskar and it talks to a local model through /v1/messages. Codex-style tools can use /v1/responses. Editors whose agents let you set an OpenAI or Anthropic base URL work the same way, and any editor with MCP can also use Toskar as tools to ask your local models and search your docs. How well an agent works depends on the model: larger coding models handle long tool lists better.
- What hardware do coding models need?
- Qwen 2.5 Coder 14B runs comfortably on a Mac with 16 GB of unified memory or a PC with 16 GB of memory; the 7B and 1.5B versions run on less. The home page's "What can my computer run?" shows your computer.
- Is temperature or max_tokens honored?
- They are accepted and ignored for now. The API docs list every known difference from OpenAI's API.
Run it for your team
Install it yourself in a few minutes, or have Yeix size the machine, set up keys and a TLS proxy, and keep it running. See also why local AI.