Changelog.

Product updates across supermemory API, MCP, plugins, SDKs, local tooling, and research.

Supermemory Console updates#

A new sidebar, smoother memory graphs, and fixes for Notion sync and container merges.

Fixes

  • Merge containers with overlapping customIds without errors. The incoming copy gets a _merge suffix.

Agents can discover the API spec at /openapi.json#

The OpenAPI spec is now served from the conventional root path on both the API host and supermemory.ai, so an agent or crawler can find it without knowing the versioned URL.

Fixes

  • Unknown routes and thrown errors now return JSON error bodies instead of plain text, matching every other API response.

Both of these serve the live spec:

https://api.supermemory.ai/openapi.json
https://supermemory.ai/openapi.json

The versioned paths still work and are unchanged:

https://api.supermemory.ai/v3/openapi
https://api.supermemory.ai/v4/openapi

Automatic memory recall and a live statusline in Claude Code#

Supermemory now pulls relevant memories into every substantive prompt on its own, and the Claude Code status bar tells you what it did: how many memories it loaded, captured, and recalled this session, and how long ago it last saved.

Improvements

  • Connect to the hosted MCP server with an existing Supermemory API key instead of going through OAuth.

Fixes

  • Documents that have a usable summary no longer save with the title Untitled Document.

The status bar rotates through these, roughly every four seconds:

6 loaded · 6 captured · 2 recalls
saved 12s ago
recalled 3m ago

Backfill historical documents with documentDate on ingest#

Set documentDate on each document so historical content lands on the date it actually happened instead of the date you uploaded it, which is what makes time-based search work on backfilled data.

Improvements

  • A new guide walks through backfilling dated documents through the batch ingestion API, with TypeScript and Python examples.

Fixes

  • Fixed an issue where SDK responses dropped fields the API returned, across 15 endpoints.

Send up to 600 documents per request to POST /v3/documents/batch, sorted oldest to newest:

{
  "documents": [
    {
      "id": "msg-2019-04-02-118",
      "content": "Decided to move the billing cutover to Q3.",
      "documentDate": "2019-04-02T09:15:00Z"
    }
  ]
}

Without documentDate a backfilled document is dated the day you uploaded it, so a question like "what did we decide in 2019" will not find it.

Build custom tools on top of Company Brain#

Use an API key from the Supermemory app to build custom tools and workflows on top of your Company Brain.

Improvements

Fixes

  • Large API document lists no longer fail while loading memory relations.

Existing scoped API keys work with v4 conversations#

Scoped API keys that can add memories can now call the v4 conversations endpoint with the same container-tag and write-access protections as memory ingestion.

Improvement

  • Scoped API keys that can add memories now work with POST /v4/conversations, so conversation ingestion can be limited to approved container tags without requiring a full-access key.
  • Existing scoped keys with memory-ingestion access inherit the conversations endpoint automatically because conversations follow the same access rules as add.

Dreaming batch retries, MCP client reconnection fix, and the Cursor install guide#

Stalled or failed Dreaming batches now retry without blocking newer batches in the same container, reconnecting an approved MCP client reuses its API key, and the Cursor installation guide uses the published integration package.

Highlights

  • Stalled or failed Dreaming batches now retry without blocking newer batches in the same container.
  • Reconnecting an approved MCP client no longer creates a new API key each time.

Fixes

  • Fixed an issue where frequently updated facts were missing from memory profiles.
  • Large Dreaming queues no longer fail during background memory generation.

Chrome extension imports from ChatGPT, Claude, Grok, and Gemini, plus smarter memory profiles#

The Chrome extension guides imports from ChatGPT, Claude, Grok, and Gemini; long memory profiles preserve recent, frequently updated facts while summarizing older context; and shareable integration links open a specific connection ready to set up.

Highlights

Fixes

Console billing status, chat space default, and MCP key fix#

The console billing page shows when a cancelled plan's access ends, organization names can be updated from web app settings, the chat input space selector defaults to Auto, and MCP clients reuse their active API key on reconnect.

Fixes

  • Fixed an issue where MCP connections created a new API key on every request instead of reusing one.

Codex plugin routes memories into project Spaces#

Codex can organize captured work into custom memory containers and keep project Spaces named from the context it sees.

Improvements

  • Memories now show a temporary title while they are still being processed.

Fixes

  • Fixed conversations ingested from plugins being attributed to the API instead of their actual source.

Smarter memory generation with timestamps, settings redesign, and Drive PDF previews#

Memory generation now uses timestamps and recent history so new information connects to related memories more accurately, the settings page is redesigned with billing in its own tab, and Google Drive PDFs preview inline in the document viewer.

Highlights

Fixes

  • Fixed an issue where Memory of the Day could link to a deleted document and open a blank page.
  • Document processing now completes even when memory generation fails for part of a batch.

Playground Auto-Learn, SMFS Docker quickstart, and plugin memory previews#

The console playground can save completed exchanges as memories with Auto-Learn, SMFS ships a Dockerfile and devcontainer quickstart with OpenAI and LangChain example agents, and memories saved by plugins and MCP render with rich previews.

Fixes

  • Fixed an issue where Pro plan users were asked to upgrade before inviting their included teammates.
  • Fixed the memory of the day card opening the list view instead of its source document.
  • Large connector syncs now import documents up to your quota instead of failing the whole batch.
  • Fixed an issue where email-change confirmation emails sometimes never arrived.

Polish

Usage-based API billing#

API billing now follows metered usage instead of relying only on fixed quotas.

Improvements

  • The file upload endpoint now accepts customId and entityContext, matching the memory add endpoint.
  • Usage analytics now report token usage per API key for tracking against plan limits.

Fixes

  • Fixed an issue where document metadata updates didn't appear in search results.

Nova redesign: new dashboard, timeline view, and connection management#

Nova gets a redesigned home dashboard, timeline view, refreshed onboarding, and expanded settings; connections move into the Add Memory modal; plus fixes for account deletion, MCP OAuth discovery, billing resilience, and the console.

Highlights

  • MCP clients can now load OAuth discovery metadata reliably.

Fixes

Mount memory as a local folder with SMFS#

SMFS mounts a Supermemory container as a local folder that shells, editors, and agents can use through ordinary file operations.

SMFS turns a Supermemory container into a folder tools already understand.

Fixes

  • Fixed PATCH document updates returning a 409 conflict against the document's own filepath.

Console .docx import, SMFS Python tool, and search fixes#

The examples repository adds research-assistant, knowledge-base, and code-sandbox templates, Console's Import tab accepts .docx files, and the SMFS repository gains an alpha Python tool for file operations and semantic memory search.

Improvements

Fixes

  • v4 search now reranks results when requested instead of silently returning the original order.

Persistent memory plugin for OpenAI Codex#

Install the Codex plugin to recall project and user memory, capture completed sessions, and explicitly search or save context.

Improvements

  • OAuth clients can now discover Supermemory through /.well-known/openid-configuration.

Connector sync cleanup, MCP project list refresh, and email polish#

Deleting an organization or account stops connector syncs and fully disconnects connections, MCP project lists refresh after a new project is used, and transactional emails use shorter layouts with plain links and one-time codes.

Fixes

  • Deleting an organization or account now stops connector syncs and fully disconnects connections.

Usage metering and Console onboarding fixes#

Organizations with overages off now stop at their included allowance, organization creation during onboarding reliably activates and continues, and document updates count only newly added content.

Fixes

  • Organizations with overages off now stop at their included allowance instead of recording billable usage.
  • Updating a document now counts only newly added content instead of re-counting unchanged text.

Test and tune memory search in the Console#

Run v3 or v4 searches, tune retrieval settings, and inspect scored results, timing, or raw JSON before integrating the API.

Improvements

  • Markdown documents now keep headings, tables, and fenced code together when split for indexing.

Fixes

  • New API keys now appear immediately in Console, and deleted scoped keys are removed completely.
  • Document processing now recovers from blocked generated output instead of failing memory generation.

Polish

  • API key previews now preserve the recognizable prefix and use a consistent masked length.

Console organization limits, memory graph, and connector fixes#

Console stops organization creation at the three-organization limit with guidance on requesting more, free organizations lose overage access after a paid plan is canceled, and the memory graph is readable in light and dark mode.

Fixes

  • Free organizations no longer keep overage access after a paid plan is canceled.
  • The API now rejects connector documents without an access token instead of ingesting their URL as text.

Organization reset and Console permission fixes#

Owners and admins can reset an organization while keeping members, billing, and the default project, CLI-created agents appear on Console's Agents page, and restricted members no longer see admin-only commands.

Improvements

  • Owners and admins can reset an organization, clearing its data and connections while keeping members, billing, and the default project.

Fixes

  • Starting User Insights twice no longer creates competing runs or clears the active report.

User Insights, memory list API, and MCP authorization fixes#

Console labels User Insights and team invitations as Scale or Enterprise features before setup, Notion connector docs clarify how documentLimit applies per sync, and User Insights now completes for larger organizations with visible processing state.

Improvements

Fixes

  • User Insights now completes for larger organizations and shows its processing state immediately.
  • The v4 memory list endpoint now returns results from every requested container tag.

User Insights: reports on user segments and content patterns#

Generate Console reports about user segments, knowledge domains, and content patterns, with an optional prompt to focus the analysis.

Fixes

  • You can now forget memories from a Console profile without getting a 404 error.
  • Usage emails no longer send the 80% warning after an organization has already reached its limit.

Video ingestion, Notion imports, and document processing fixes#

API video ingestion no longer fails while extracting transcripts and descriptions, Notion imports explain unshared or linked databases and how to continue, and documents with large memory batches finish processing.

Fixes

  • API video ingestion no longer fails while extracting transcripts and descriptions.
  • Documents that produce large memory batches now finish processing instead of failing during memory generation.
  • Documents no longer fail when a generated memory references another memory that was filtered out.

Nova onboarding, API search, and OpenClaw plugin fixes#

Empty Nova spaces can switch to All Spaces to find memories stored elsewhere, raw base64 images are recognized correctly during summarization, and OpenClaw recognizes Supermemory as its active memory backend.

Fixes

  • Image content sent as raw base64 is now recognized as JPEG, PNG, GIF, or WebP instead of failing during summarization.
  • Metadata filters containing null characters no longer cause search requests to fail.

Console improvements: API key flow and request analytics#

Console's API key flow gets a guided side panel for expiry and scope with the key shown once, and the Requests page summarizes request volume, success rate, average search latency, and request types with per-key filtering.

Improvements

Document ingestion reliability fixes and Nova improvements#

Document ingestion can run up to three targeted searches to retrieve related memories as context, Nova explains region-unavailable models and offers alternatives, and the open-source LLM Bridge library supports OpenAI Responses and cross-provider streaming.

Improvements

  • During document ingestion, Supermemory can now generate up to three targeted searches to retrieve related existing memories as context.

Fixes

  • Documents no longer remain permanently stuck in processing after a workflow stops; Supermemory retries them once and marks a second failure clearly.
  • When scheduling a batch document upload fails, only the affected batch is marked failed instead of every queued document.
  • API and MCP quota errors now name the limit reached and explain how to continue with overage.
  • Truncated model output during ingestion no longer discards every generated memory when only relation data is incomplete.
  • Updating a document with the same custom ID no longer creates duplicate memories from facts already extracted.
  • Web links that return an authentication, paywall, or rate-limit response during metadata lookup no longer fail the entire import.
  • Search requests with an empty query or an out-of-range limit now return a clear client error instead of a server error.

Memory expiration fields, plus Nova, MCP, and plugin fixes#

New v4 memories can carry an expiration time, a reason for forgetting, and document or event dates, sign-ups with unsupported email domains get a clear rejection, and usage limits account for full request size.

Improvements

Fixes

  • Sign-up attempts with an unsupported email domain now return a clear rejection instead of a generic server error.
  • Usage limits now account for the full size of each request and allow continued processing when organization overage is enabled.
  • Nova no longer intermittently shows an Upgrade to Pro prompt on plugin and connection pages for users with an active paid plan.
  • MCP connections authenticated with an API key no longer fail with a missing-organization error.

Console documents page, scoped API keys, and connector fixes#

Scoped API keys can grant access to multiple container tags, Console's organization picker shows each organization's plan and usage totals, and empty Nova spaces offer direct actions to start saving content.

Improvements

Fixes

  • Console's API key list no longer crashes when an existing key has no metadata.
  • Deleting or merging container tags no longer reports a failure after the operation has already succeeded.
  • Google and Microsoft connections now keep syncing when new sign-in credentials replace the old ones.
  • Updating only a document's metadata no longer reprocesses unchanged content.

Console billing updates and Documents API audio transcription#

Paid Console users can open the billing portal from the Billing page, the Requests page gets adjustable page sizes, connector setup lets you choose a document limit, and the Documents API adds audio transcription.

Improvements

Fixes

  • Container tags containing colons work again in API requests.
  • The Documents API can now retry a failed document with the same custom ID instead of returning the failed record again.

Microsoft Agent Framework integration, Console and connector fixes#

Microsoft Agent Framework gains a Supermemory integration with automatic recall, optional conversation storage, and explicit memory tools, Nova adds multi-select document deletion, and Console admins can copy pending invitation links.

Improvements

Fixes

  • Container tags with spaces or symbols no longer prevent documents and memories from being indexed.

Plugin sign-in approval, Console and Nova fixes#

Plugin sign-in now shows an approval page before returning an API key to the requesting app, OpenCode gains a /supermemory-logout command, OpenClaw builds memory from user messages while treating agent replies as context, and plugin sign-in works for users with multiple organizations.

Improvements

  • Plugin sign-in now shows an approval page before returning an API key to the requesting app.

Fixes

  • Container tags containing URL-sensitive characters can now be viewed, updated, and deleted correctly in Console.
  • Console now shows clear not-found states for missing documents and container tags.

MemoryBench providers, Claude skills, and container tag filtering#

MemoryBench gains a Claude skill for scaffolding provider adapters and running benchmarks, new filesystem-memory and hybrid-RAG providers for comparison, and the Supermemory repository gets a Claude skill with setup guidance and TypeScript and Python examples.

OpenClaw custom memory containers, plus API and Nova fixes#

OpenClaw can route store, search, forget, and profile operations to custom memory containers, document listing matches any selected container tag, and throttled API keys get a proper rate-limit error.

Fixes

  • Document listing now returns documents matching any selected container tag instead of requiring every tag.
  • The API now returns a rate-limit error for throttled keys instead of reporting them as unauthorized.
  • v4 profile searches and search requests now record search-query usage correctly.
  • Large documents now finish processing instead of failing midway or timing out during summarization.

OpenCode container tags, MCP forget fix, and Nova polish#

OpenCode can use exact user and project container tags reusable across teammates and machines, plus fixes for saving captured conversation transcripts and MCP's forget action.

Fixes

  • Short notes and questions can now finish processing even when they do not need a generated summary.

API fixes: safer array filters and organization scoping#

Array filters in search, memories, and connections now safely handle user-provided values, and related results stay within the active organization.

Fixes

  • Array filters in search, memories, and connections now safely handle user-provided values, and related results stay within the active organization.

Search deduplication and container tag scoped API keys#

Search results deduplicate identical content down to the highest-scoring match, Console can create API keys scoped to a single container tag, and owners and admins can delete a container tag with its memories and documents through the API.

Improvements

Fixes

Console plugins page and automatic document ingestion retries#

Console gains a dedicated Plugins page for browsing and managing plugins, and documents stuck in the queue now retry automatically up to three times before being marked failed.

Fixes

  • Documents that remain queued now retry automatically up to three times before being marked failed.

Notion page properties in document metadata and append fixes#

Notion connections expose page properties like status, tags, dates, and assignees as document metadata for filtering and queries, and appending to a document now retains the full combined content with its total token count.

Improvements

Fixes

  • Appending to an existing document now retains the full combined content and reports its total token count.

AI SDK strict mode, MCP recall, and browser extension fixes#

AI SDK memory tools support strict schema validation for models that require every parameter, MCP recall results include the memory text next to each match percentage, and pages saved by the browser extension keep their original source link.

Fixes

  • Pages saved by the browser extension now appear as webpages with their original source link instead of plain-text documents.

MCP updates: project listing and API key reuse fix#

The MCP server can list your projects and use their names when saving or recalling memories, and reconnecting an approved MCP client reuses its active API key instead of creating another one.

Fixes

  • Reconnecting an approved MCP client now reuses its active API key instead of creating another one.

PDF import reliability and console organization fixes#

PDF imports retry failed pages with a fallback extractor without truncating long documents, and Console no longer shows accepted invitations as pending or overflows on long organization IDs.

Fixes

  • PDF imports now retry failed pages with a fallback extractor and avoid truncating long documents.
  • Fixed an issue that caused some multi-chunk documents to fail during processing.

Console usage stats, AI SDK prompt templates, and OpenCode updates#

Delete a consumer organization and its stored data from Profile settings, control how retrieved memories are formatted in AI SDK prompts, and see token and search-query usage in Console for the current and previous billing periods.

Improvements

  • You can delete the active consumer organization and its stored data from Profile settings.

Fixes

  • Updating a document now processes and bills only newly added content instead of unchanged chunks.

MemoryBench parallel execution and MCP connection reliability#

MemoryBench can run benchmark phases concurrently with separate concurrency controls per phase, and MCP and streaming connections retry brief server overloads before returning a temporary-unavailable error.

Fixes

  • MCP and streaming connections now retry brief server overloads before returning a temporary-unavailable error.

MCP memory limits and Google Drive connector fixes#

Memories added through MCP now count against the memory allowance instead of the API token limit, and the Add Memory connection picker opens Google Drive instead of returning a disabled-connector error.

Fixes

  • Memories added through MCP now count against the memory allowance instead of the API token limit.
YOUR PRIVACY

Change or withdraw any time via Cookie settings in the footer. Read our cookie notice.