One chat panel. Any provider. Your keys, your models.
Tempr is a bring-your-own-key (BYOK) extension: a rich in-IDE chat, inline completion, and agent toolset inside Visual Studio, VS Code, and JetBrains IDEs (IntelliJ IDEA, PyCharm, WebStorm, Rider, and more) that works with first-party providers (OpenAI, Anthropic, Google) and aggregators (OpenRouter and others). Add your own API key and pick whichever model fits the task — nothing routed through a shared pool.
Choose a plan See it in action
One account. Four ways to build.
Every Tempr product is bring-your-own-key and shares one login and one Portal — pick where you want the agent to live, or mix and match. A fourth, Tempr Code, is on the way.
Tempr Chat
The full agent toolset — chat, inline completion, personas, and MCP servers — right inside Visual Studio, VS Code, and JetBrains IDEs.
Explore ChatTempr CLI
The same agent in your terminal: one-shot prompts, an interactive REPL, and piped headless mode for scripts and automation.
Explore CLITempr Code Soon
The agent in its own desktop app: parallel chats on isolated worktrees, a built-in terminal, and a diff you approve before anything lands.
Take a lookTempr Gateway
An OpenAI-compatible API in front of your own provider keys — virtual keys, request logs, analytics, and quotas.
Explore GatewayEverything you need, right inside the IDE
No more tab-switching to a browser. Chat, complete, and refactor without ever leaving your IDE.
Bring your own key(s)
Connect directly with first-party providers (OpenAI, Anthropic, Google, xAI) or an aggregator (OpenRouter and others) using your own API key — no shared usage pool, no vendor lock-in. Required on every plan.
Rich in-IDE chat
Rich Markdown and code block rendering, one-click copy/insert at cursor, slash commands (/explain, /fix, /tests, /doc), and full chat history you can revisit anytime.
Agent mode, your way
A real tool-calling loop that reads, searches, and edits your codebase and runs whitelisted commands — choose Interactive (approve every change), Bypass (auto-apply edits), or Autopilot.
Named agent personas
Type @debugger, @test, @security, @review, @docs, @refactor, @modernize, @profiler, @git, or @plan for a specialist prompt and tool scope. Standard unlocks the full built-in set; Pro adds custom personas you define yourself.
Multi-agent delegation
The agent can hand focused sub-tasks off to specialist personas mid-turn — e.g. delegating root-cause analysis to @debugger while it keeps working — running several in parallel when it helps.
Real semantic code search
Embedding-based search over your solution finds conceptually related code even when the wording doesn't match — not just grep with extra steps.
Debugger-aware
The agent can inspect the live call stack and locals and evaluate expressions during a debugging session, for context-aware help grounded in your actual running process.
Test & solution aware
Lists and runs your tests with structured pass/fail results, reads live compiler diagnostics, and understands your solution's project references and call graph.
Inline completion
Ghost-text style inline completions that suggest the next line as you type, tuned per-scenario to the model you prefer.
Model Context Protocol (MCP)
Connect external MCP servers to extend the agent with your own tools and data sources.
Resilient agent runs, synced everywhere
The agent loop runs server-side, so it survives a dropped connection and picks back up mid-turn instead of losing progress — while tool execution (reading, editing, and running commands) still happens locally in your IDE. Chat sessions, tool preferences, and custom personas sync to the cloud too (Standard+) so your setup follows you across machines.
Simple, tiered licensing
Standard, Pro, and Teams all unlock the full model catalog — Pro and Teams get new models immediately, Standard about 7 days later — plus deeper agent features like custom personas, multiple saved sessions, and higher agent-turn limits. Teams adds centralized billing, seat management, and shared team personas/config on top of everything in Pro. Managed via Stripe subscriptions with automatic license-key issuance.
Bring a key from any of these
Around 30 providers on the actual server-side allowlist Tempr enforces today — first-party APIs, inference hosts, clouds, and aggregators. A few of the most popular:
See it in action
The Tempr chat panel inside Visual Studio and VS Code, and the Tempr Portal for managing licenses and usage.
How it works
1. Install the extension
Install the Tempr extension into Visual Studio from the Visual Studio Marketplace, into VS Code from the Visual Studio Code Marketplace, or into JetBrains IDEs (IntelliJ IDEA, PyCharm, WebStorm, Rider, and other IntelliJ-Platform IDEs) from the JetBrains Marketplace.
2. Activate a license
Subscribe to Standard or Pro — license keys are emailed automatically after checkout, and activate in seconds inside the extension.
3. Add your API key(s)
Bring your own key for a first-party provider (OpenAI, Anthropic, Google) or an aggregator (OpenRouter and others) — stored encrypted at rest, never in plain text.
4. Chat with any model
Open the tool window, pick a model from the dropdown, and start chatting, completing, or delegating to an agent.
Which Tempr do I need?
Probably more than one — they share a single login, and the CLI comes free with every Chat plan.
| You want to… | Chat | CLI | Gateway |
|---|---|---|---|
| Chat, complete, and refactor inside Visual Studio, VS Code, or JetBrains IDEs | ✔️ | — | — |
| Run the agent from a terminal or inside scripts and automation | — | ✔️ | — |
| Call an OpenAI-compatible API from your own app or backend | — | — | ✔️ |
| Virtual keys, request logs, analytics, and spend visibility for API traffic | — | — | ✔️ |
| Bring your own provider key — any product, any plan | ✔️ | ✔️ | ✔️ |
What developers are saying
Illustrative quotes for preview purposes — replace with real customer feedback before launch.
"I like that it's BYOK — I use my own OpenAI key for day-to-day work and my OpenRouter key when I want to try something like DeepSeek."
— Backend developer, fintech startup
"The agent personas are a great shortcut — @debugger genuinely saves me time chasing down flaky tests."
— Senior engineer, SaaS platform
"No usage pool, no rate-limit surprises — it's just my own key talking straight to the provider. Exactly what I wanted."
— Independent contractor
Chat capabilities, plan by plan
Everything in Standard, Pro, and Teams, side by side — the CLI rides with every plan.
| Capability | Standard | Pro | Teams |
|---|---|---|---|
| Requires your own provider API key(s) | ✔️ | ✔️ | ✔️ |
| Tempr CLI on the same license | ✔️ | ✔️ | ✔️ |
| Model catalog | Full (new models after ~7 days) | Full (immediate) | Full (immediate) |
| Chat & inline completion | ✔️ | ✔️ | ✔️ |
| Agent tooling & built-in personas | ✔️ | ✔️ | ✔️ |
| Custom personas | — | ✔️ | ✔️, shared across the team |
| Custom models (your own OpenAI-compatible endpoints) | — | ✔️ | ✔️, org-managed |
| Multi-agent delegation | ✔️ | ✔️ | ✔️ |
| External MCP servers (Model Context Protocol) | ✔️ | ✔️ | ✔️ |
| Real semantic code search | ✔️ | ✔️ | ✔️ |
| Cloud sync (sessions, tools, personas) | ✔️ | ✔️ | ✔️, team-wide |
| Saved chat sessions | 1 | Multiple | Multiple |
| Agent-turn step limit | 50 | 100 | 100 |
| Delegated sub-agent calls per turn | 3 | 8 | 8 |
| Parallel sub-agent delegation | 2 at a time | 8 at a time | 8 at a time |
| Priority feature requests | — | ✔️ | ✔️ |
| Centralized billing & seat management | — | — | ✔️ |
| Admin controls & usage visibility | — | — | ✔️ |
| SSO (SAML/OIDC) | — | — | ✔️ |
| Org-managed virtual keys (pooled provider keys, per-member limits) | — | — | ✔️ |
Teams starts at $35/seat/month with a 10-seat minimum. Reach out via the contact form and we'll follow up with a quote and a signup link sized to your team.
Chat FAQ
The questions specific to the IDE extension. Account, BYOK, privacy, and billing questions live on the general FAQ page.
Contact us
Questions about pricing, a Teams plan, or something not working as expected — send a message and we reply to every one.
Prefer email? Reach us directly at contact@temprhq.io.
Also on your account
The same agent, models, and keys you set up here work everywhere — one login.