See where your AI spend goes.
Follow AI spend from your organization down to a session, set team budgets, and manage Enterprise devices.
Releases
Gemini on Vertex AI
Gemini calls on Vertex AI can now go through condense for context management, on your own Google account and with the same request format.
Check your Claude plan limits with dense usage
dense usage shows how much of your Claude plan you've used, next to an estimate of what it would be without condense. Claude subscriptions only.
Check your savings with dense info
dense info shows your condense account, lifetime savings, and the context and spend of the session you're in, from the terminal or inside your agent.
Bedrock and Vertex AI routes
Teams on AWS or Google Cloud can point their Bedrock and Vertex AI calls at condense to manage their context, on their own cloud account.
Condense Lite
When your weekly pool runs out, Condense Lite keeps managing your context and saves up to 20% of each prompt. It's on every plan, including Free, at no extra charge.
Helene 1.1
Beats bear-2 by 7.1 accuracy points and kompress-v2 by 12.5 at matched compression on the public Lost in Compression benchmark.
Introducing Helene 1
On CoQA, a model answering from its compacted context is more accurate than one reading the full transcript, with almost a third fewer tokens.
Built for teams. Building for the enterprise.
Organizations are live: one workspace, a shared credit pool, a per member budget, and a spend cap you set.
Adeline V1 ships. The API ships with it.
Adeline 1, our compaction model, goes live, and the proxy now speaks both the OpenAI and Anthropic SDKs.
Hello, condense.chat
Faster context compaction for Claude Code and the Claude Agent SDK, in beta.