Yiorgis Gozadinos 7f7223e0ac

Replaces gpt-oss on ModelConfig, qa.model and processing.title_model, and
ministral-3 on the picture-description model. qa.model.vision follows the
model and is now true.

enable_thinking was gated on the gpt-oss name, so it did nothing for
qwen3.8. With title_model's max_tokens of 100 the reasoning consumed the
whole budget and title generation returned an empty string. The mapping
now applies to any ollama model via reasoning_effort(): false sends
"none", true sends "high". Measured on qwen3.8:27b-mlx, "low" does not
disable thinking and "none" does; gpt-oss is the inverse, its template
has no "none" level, so it keeps "low".

Picture description bypasses get_model -- docling posts the request
itself from a params dict -- so the flag was inert on that path too.
vlm_api_params() carries reasoning_effort into both converters' request
bodies. At max_tokens 200 the description survived either way, but the
switch cut completion tokens from 141 to 45.

test_search_tool_skips_binary_content_when_qa_model_is_text_only asserted
the vision default rather than setting it; it now configures vision=False
itself.

docs/benchmarks.md keeps ministral-3: those are recorded measurements.

2026-09-04 12:36:36 +03:00

2.3 KiB

Raw Blame History

Quickstart

Install haiku.rag, index a document, and chat with it.

Install

uv pip install haiku.rag

You also need Ollama for the default embedding and answering models:

ollama pull qwen3-embedding:4b
ollama pull qwen3.8

!!! note "Prefer OpenAI?" Drop this into a haiku.rag.yaml next to where you'll run the CLI:

```yaml
embeddings:
  model:
    provider: openai
    name: text-embedding-3-small
    vector_dim: 1536

qa:
  model:
    provider: openai
    name: gpt-4o-mini
```

Then `export OPENAI_API_KEY="sk-..."` and continue with the rest of this page. Any provider Pydantic AI supports works the same way. See [Providers](configuration/providers.md).

Initialize

haiku-rag init

This creates a LanceDB database in your platform's user directory. Pass --db to any subcommand to use a different path:

haiku-rag init --db /tmp/test.lancedb

Add a document

Add a file, a URL, or a whole folder:

haiku-rag add-src https://arxiv.org/pdf/2408.09134
haiku-rag add-src ~/Documents/papers/

Or paste text inline:

haiku-rag add "Yiorgis wrote haiku.rag in 2025."

Each add-src call converts the file with Docling, splits it into chunks, embeds them, and writes everything to LanceDB. Run haiku-rag list to see what you've added, haiku-rag info for a database summary.

Chat

haiku-rag chat

Ask a question. The agent searches your documents, expands context around the hits, and answers with citations pointing back to the source page and section. Citations are expandable, with visual grounding so you can see the chunk highlighted on the original page. Follow-ups continue within the same session. Start a new session when you switch topics.

You can also ask a single question directly from the CLI without launching the TUI:

haiku-rag ask "Who wrote haiku.rag?"

Where to go next

Chat: sessions, citations, and the full TUI.
CLI reference: every command.
Python API: use haiku.rag in your own code.
Capabilities: native RAG and analysis components used by the client.
Tuning: better retrieval.
Configuration: every setting.

2.3 KiB Raw Blame History