{"slug": "what-s-new-in-hindsight-1-2-0", "title": "What's new in Hindsight 1.2.0", "summary": "Hindsight 1.2.0 adds support for the Hindsight models — the hs-decide-1-flash decision model, hs-rerank-1 reranker and hs-embed-1 embedding model — and makes hs-rerank-1 the default reranker in place of cross-encoder/ms-marco-MiniLM-L-6-v2, the release notes state. The release also adds per-request and per-bank prune and dedup controls for decision-model recall, routes reflect's tool calls through a response schema so they work on OpenAI-compatible servers such as vLLM that ignore forced tool choice, and ships reliability fixes across retain, consolidation, reflect, refresh and the worker. The release notes tell everyone on 1.1.x to upgrade.", "body_md": "# What's new in Hindsight 1.2.0\n\nSupport for the Hindsight models, a new default reranker, prune and dedup controls for decision-model recall, images on knowledge pages, a better LLM request view, and a long round of reliability fixes\n\nHindsight 1.2.0 adds support for the Hindsight models, makes `hs-rerank-1` the default reranker, gives recall finer control over what the decision model keeps, shows the images behind knowledge pages, and makes the LLM request view easier to read. Tool calls now go through a response schema on every provider, which makes reflect far more reliable on self-hosted models. Most of the release is reliability work across retain, consolidation, reflect, refresh and the worker. **Everyone on 1.1.x should upgrade** — see the [changelog](https://hindsight.vectorize.io/changelog) for the full list.\n\n- [**Hindsight models**](#hindsight-models) : A decision model, a reranker and an embedding model built for agent memory;`hs-rerank-1` is the new default reranker.\n- [**Decision-model recall: prune and dedup**](#decision-model-recall-prune-and-dedup) : Drop irrelevant and duplicate memories, per request or per bank.\n- [**Tool calls through a response schema**](#tool-calls-through-a-response-schema) : Reflect works on servers that ignore forced tool choice.\n- [**Images on knowledge pages**](#images-on-knowledge-pages) : See the images behind a page's evidence.\n- [**A better LLM request view**](#a-better-llm-request-view) : Retention, an hourly chart, and a split by model or operation.\n- [**More settings**](#more-settings) : Vision call settings, embedding extra body, a default time zone.\n- [**Helm**](#helm) : Optional API command, startup probe, strategy and token mounting.\n\n## Hindsight models\n\n1.2.0 supports the Hindsight models: `hs-decide-1-flash`, a decision model that takes over the small judgments in consolidation, reflect and knowledge pages, plus `hs-rerank-1` and `hs-embed-1`. They're off until you set them up: run `hindsight-admin cloud login` and set `HINDSIGHT_API_DECISION_MODEL_PROVIDER=hindsight`. See [Decision Models](https://hindsight.vectorize.io/developer/decision-models) for every setting.\n\n**The default reranker changes.** The local and ONNX rerankers now use `hindsight-ai/hs-rerank-1`, multilingual (eight languages) and trained for Hindsight recall, instead of `cross-encoder/ms-marco-MiniLM-L-6-v2`. It downloads on first start. To keep the old model on the local provider, set `HINDSIGHT_API_RERANKER_LOCAL_MODEL=cross-encoder/ms-marco-MiniLM-L-6-v2`.\n\nThe `typesafe` reranker provider is deprecated in favour of the shared decision model.\n\n## Decision-model recall: prune and dedup\n\nWhen the decision model reranks recall, it can now also drop memories that don't help answer the query (`prune`) and memories that repeat one already returned (` dedup`). Turn them on per request (HTTP and MCP recall), or per bank with `HINDSIGHT_API_RECALL_DECISION_MODEL_PRUNE` and `HINDSIGHT_API_RECALL_DECISION_MODEL_DEDUP`. The decision model now also sees each memory's date, so time-sensitive questions rank better.\n\n## Tool calls through a response schema\n\nReflect's tool calls now go through a response schema instead of the provider's native tool calling. Some OpenAI-compatible servers (for example vLLM with some open models) accept tools but don't honour a forced tool choice, so reflect's forced steps were often skipped. Those servers do honour a response schema, so reflect now behaves the same on every provider.\n\n## Images on knowledge pages\n\nWhen a knowledge page's evidence comes from retained images, the page now shows them next to that evidence in the control plane.\n\n## A better LLM request view\n\nThe LLM request view says how many days of history are kept, adds an hourly chart for the last 24 hours, and splits usage by model or by operation. It now also records embedding, rerank and decision-model calls, not only LLM calls. Set `HINDSIGHT_API_LLM_TRACE_DELETE_WITH_DOCUMENT=true` to delete a document's request history when the document is deleted; deleting a bank always deletes its history.\n\n## More settings\n\n- **Vision calls** get their own reasoning effort, timeout, retries and extra body (`HINDSIGHT_API_VLM_*` ).\n- **OpenAI-compatible embeddings** can send an extra request body (`HINDSIGHT_API_EMBEDDINGS_OPENAI_EXTRA_BODY` ).\n- **Default time zone** (`HINDSIGHT_API_DEFAULT_TIMEZONE` ): retain timestamps without a UTC offset are read in this zone.\n\n## Helm\n\nThe chart adds an optional API `command`, a `startupProbe`, a deployment `strategy`, and control over `automountServiceAccountToken`. Docker and Helm give the API 60 seconds to finish in-flight work on shutdown.\n\n**Breaking changes:** no provider is sent native tools any more, so an OpenAI-compatible server must support a response schema (structured output). `HINDSIGHT_API_LLM_SUPPORTS_MAX_ITEMS`, `HINDSIGHT_API_LLM_STRUCTURED_OUTPUT_FORCED_TOOL` and `HINDSIGHT_API_REFLECT_PROMPT_CACHE_ENABLED` are removed; remove them from your configuration. On the LiteLLM-backed providers (`litellm`, `litellmrouter`, `bedrock`), set `HINDSIGHT_API_LLM_STRUCTURED_OUTPUT_IN_PROMPT=true` to give the schema in the prompt instead.", "url": "https://wpnews.pro/news/what-s-new-in-hindsight-1-2-0", "canonical_source": "https://hindsight.vectorize.io/blog/2026/11/03/version-1-2-0", "published_at": "2026-11-03 00:00:00+00:00", "updated_at": "2026-11-03 17:24:14.702739+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "large-language-models", "ai-infrastructure"], "entities": ["Hindsight", "hs-decide-1-flash", "hs-rerank-1", "hs-embed-1", "cross-encoder/ms-marco-MiniLM-L-6-v2", "vLLM"], "also_reported_by": [{"url": "https://hindsight.vectorize.io/blog/2026/10/21/version-1-1-0", "source": "hindsight.vectorize.io", "title": "What's new in Hindsight 1.1.0", "published_at": "2026-10-21T00:00:00+00:00"}], "alternates": {"html": "https://wpnews.pro/news/what-s-new-in-hindsight-1-2-0", "markdown": "https://wpnews.pro/news/what-s-new-in-hindsight-1-2-0.md", "text": "https://wpnews.pro/news/what-s-new-in-hindsight-1-2-0.txt", "jsonld": "https://wpnews.pro/news/what-s-new-in-hindsight-1-2-0.jsonld"}}