Vision, voix et multimodal

Gives a text-only model image understanding via a vision-language model, exposed as a `describe_image` tool.

Plugin DeepSeek Harness qui relie les images de session à des API de vision interchangeables, tout en gardant DeepSeek comme modèle principal.

Passerelle de vision native au fournisseur LLM : les images collées dans le chat sont décrites par un modèle de vision (Qwen3-VL via pi-ai/llama.cpp), et la description textuelle est transmise à DeepSeek texte seul pour la réponse — l'admission, le routage et la compaction des images passent tous par les mécanismes natifs du harness, avec un cache de descriptions LRU et une reprise sur erreur 503.

Les images jointes au composeur sont transcrites en texte par un modèle de vision compatible OpenAI avant d'atteindre les modèles DeepSeek texte seul.

Saisie vocale au micro pour le composeur : transcription en direct via la Web Speech API du navigateur, déduplication/poursuite automatique, ponctuation intelligente, réglages de langue et d'envoi automatique.

Voice pair: free edge-tts neural speech synthesis + OpenAI-compatible ASR transcription.

Voice input for the Web UI: a mic button that drives local VocoType offline speech recognition and auto-inserts recognized text into the composer (auto-deploy, dedupe, continuous dictation).

Mic button in the composer tool row: Web Speech API speech-to-text (Chrome/Edge), language switching, and optional auto-send, zero dependencies.

Universal image generation for DeepSeek Harness: auto-discovers image models from any OpenAI-compatible endpoint (SenseNova, StepFun, Agnes, Qwen, Flux, SD, Imagen and more), with agent tools and REST API.

Native vision capability extension, using either Zhipu (free) or Qwen-VL (local).

Renders pasted/uploaded images inline in the DSH Web chat and gives text-only models vision: the model-invokable visual_review tool calls any OpenAI-compatible multimodal API first, falling back to a local Qwen3-VL worker.

Vision-augmented DeepSeek adapter: a vision-capable model describes image input, then a text-only DeepSeek model reasons over the description.

Full vision-capability bundle for DeepSeek Harness: a vision_understand tool (OpenAI-compatible vision APIs, free Zhipu GLM-4V-Flash by default) plus paste/drag-and-drop/button entry points for image recognition.

Local OCR for attached images via Tesseract: only the recognized text is sent to the model, never the image bytes; vision passthrough is opt-in.

GPT Image 2 `image_gen` with Codex subscription OAuth by default or explicit API-key mode: developing card, up to three live API partials, durable attachment replay/lightbox/download, text-only model output, and bounded credential-safe requests.

Vision bridge for text-only DeepSeek routes that analyzes attached and local images through configurable OpenAI Responses, Chat Completions, or Anthropic Messages endpoints while leaving image-capable routes native.

Combine text, vision, and image-generation APIs into one Mix model with automatic routing: text-only requests go to the chat model, user images and agent screenshots go to the vision model, follow-ups keep using the same session image, and agents can generate or edit images with session-scoped call history.

Free vision bridge for text-only models: image understanding, OCR, UI and debug analysis via free-tier providers (Qwen3-VL-Flash, Doubao, DeepSeek-OCR) with a settings GUI.

Agent-initiated voice calls: `offer_call` rings the human (接听/拒接/稍后再说); accepted calls synthesize and play locally via CrispASR + Qwen3-TTS (9 speakers, 2 Chinese dialects), rejected calls return the decision to the agent.

Free voice closed loop for the Web UI: browser SpeechRecognition mic input with live interim results plus read-aloud speaker buttons and auto-read for assistant replies, zero configuration and no API key.

Speech-to-text voice input for the web UI: a mic button in the composer transcribes speech into the draft via the browser Web Speech API (zero-config) or an OpenAI-compatible Whisper API (OpenAI / Groq), with a selectable model and language in Settings.

Chat image-attachment bridge with a `view_image` tool for any OpenAI-compatible VLM (local Ollama or cloud): pasted/dropped images become `view_image` path markers before reaching text-only DeepSeek models.

`describe_image` tool: a vision bridge that sends images to mimo-v2.5 through the opencode Zen API (credential `OPENCODE_GO_API_KEY`, free route first with paid fallback) and returns text descriptions for text-only models, with native passthrough and ImageMagick transcoding of SVG/TIFF/HEIC formats.

Adds an image_describe tool backed by the DashScope OpenAI-compatible vision API; on text-only sessions, pasted images are stored as local paths for the model and rendered inline in the chat transcript.