Claude Opens Message Threads in the Messages API Public Beta
Anthropic's public beta for Message Threads lets developers send only new messages to an existing Messages API conversation instead of resending its full history. Threads are held encrypted in memory for five minutes after the last request. The feature does not reduce token usage, and Anthropic reports a median time-to-first-token improvement for long-context Claude Code requests.
Anthropic has opened Message Threads as a public beta in the Messages API. A client can continue a conversation by sending only new messages and the ID of the previous response, rather than resending the full history on every turn. (Anthropic's announcement)
How Message Threads Work
The first request includes the conversation and starts a thread with the thread parameter. Later requests include only new messages and the previous response ID. Anthropic holds each thread in encrypted server memory between turns and does not write it to disk. (Anthropic's announcement)
A thread expires five minutes after its last request. When that happens, the client resends the full conversation to start a new thread. (Anthropic's announcement)
Latency and Token Usage
Anthropic reports that Claude Code requests with 500k to 1M tokens saw about a 35% drop in median time to first token (TTFT) with message threads. (Anthropic's announcement)
Sending smaller requests does not cut token usage: Anthropic says Claude still reads the full conversation on every turn, so token usage stays the same. The client sends less repeated data, while Claude can start responding sooner. (Anthropic's announcement)
Availability and Best-Fit Workflows
The public beta is available on the native Claude Platform, Claude Platform on AWS, Claude in Microsoft Foundry, and Claude Models on Google Cloud (Anthropic managed). Standard Claude token rates and existing rate limits apply. (Anthropic's announcement)
Anthropic highlights long-running coding agents, computer-use agents, and document- or image-heavy workflows as good fits when latency matters. The feature is aimed at long-running sessions where conversation history grows across repeated turns. (Anthropic's announcement)
Read next
Related coverage- Claude
Claude API Request Limit: 128 MB for Messages and Token Counting
Anthropic's Claude Platform docs list a 128 MB maximum request size for Messages and Token Counting on the first-party API. Cloud-platform limits are smaller and vary by provider: 20 MB on Bedrock, 30 MB on Google-managed Claude, and up to 32 MB on Anthropic-managed cloud services. Oversized first-party requests return 413 `request_too_large`.
- Anthropic
Anthropic Resumes Billing for Claude Refusals Before Output in Three Safeguard Categories
Anthropic resumed billing for Claude requests refused before any output when its safeguards classify them as biology, frontier LLM development, or reasoning extraction. The charge uses the model's normal rates across platforms. Other pre-output refusals remain unbilled, but still count against rate limits.
- Claude
Claude Max, Ultra, and Team Plans Now Include Free Monthly API Credits: $100, $200, or Up to $500
Claude Max, Ultra, and Team plans include monthly Claude API credits at no extra cost: $100 on Max, $200 on Ultra (formerly Max 20x), and up to $500 pooled on Team. You claim them in claude.ai billing settings by linking one Console organization. They cover the API, Managed Agents, and the Agent SDK, not interactive Claude Code.
- Claude Code
Claude Code Mods Let Developers Customize the CLI and Desktop App
Anthropic introduced Claude Code mods as a plugin-based way to customize how Claude Code behaves and looks. Small JavaScript or TypeScript modules can add interface elements, change tool calls, or replace built-in features. Mods run with Claude Code's access to your machine, so review their source before installing.
Sources
- Anthropic, "Make long-running agents faster with message threads" (November 2, 2026)
Frequently Asked Questions
What are Claude API message threads?
They let a client continue a Messages API conversation by sending only new messages and the ID of the previous response, rather than resending the full conversation history.
How long does a message thread stay active?
Anthropic keeps thread state in encrypted memory for five minutes after the last request. After that, the client can resend the full conversation to start a new thread.
Do message threads reduce Claude token usage?
No. Anthropic says Claude still reads the full conversation on each turn, so token usage stays the same even though each request is smaller.