# DeepSeek-V4.1 text and vision encoding `encoding.py` is the standalone prompt-format reference for DeepSeek-V4.1. It supports multi-turn conversations, tool calls, thinking modes, numeric reasoning effort, mid-conversation system messages, and interleaved image content blocks, without importing the inference implementation. ## V4.1 changes relative to V4 Three prompt-format changes distinguish V4.1 from V4: 1. **DSML tag names use a leading space.** Tool calls are wrapped in `<|DSML| calls>` blocks with `<|DSML| invoke>` / `<|DSML| parameter>` tags (note the space before `calls`, `invoke`, and `parameter`). The V4 format used `<|DSML|tool_calls>` without a space. 2. **Reasoning effort is a numeric budget (1–100).** The effort prefix is rendered as `Reasoning Effort: {budget} (range 1-100, ...)` rather than the verbose natural-language descriptions used in V4. String aliases map as follows: `"low"` → 50, `"high"` → 75, `"max"` → 100. The default is `"high"` (75). The effort prefix is only rendered in `thinking_mode="thinking"` and only at the beginning of the conversation (index 0). 3. **Mid-conversation system messages** are supported via the `<|System|>` token. A mid-conversation system message behaves like a user message for the purpose of appending the assistant generation header. ## Quick start ```python from encoding import encode_messages, parse_message_from_completion_text # Text-only conversation messages = [ {"role": "system", "content": "You are a helpful assistant."}, {"role": "user", "content": "What is 2+2?"}, ] prompt, media = encode_messages( messages, thinking_mode="thinking", reasoning_effort=75, # integer 1–100, or "low"/"high"/"max" return_multi_modal_data=True, ) # prompt: # '<|begin▁of▁sentence|><|System|>Reasoning Effort: 75 (range 1-100, the higher the # value, the more thorough the reasoning)\n\nYou are a helpful assistant. # <|User|>What is 2+2?<|Assistant|>' # Parse model output back to a structured message completion = "Simple arithmetic.2 + 2 = 4.<|end▁of▁sentence|>" parsed = parse_message_from_completion_text(completion, thinking_mode="thinking") # => {"role": "assistant", "reasoning_content": "Simple arithmetic.", # "content": "2 + 2 = 4.", "tool_calls": []} ``` > **Note:** `parse_message_from_completion_text` is designed to handle > well-formatted model output only. It does not attempt to correct or recover > from malformed output that the model might occasionally generate. For > production use, additional error handling is recommended. ## OpenAI-style messages ```python from encoding import encode_messages messages = [{ "role": "user", "content": [ {"type": "text", "text": "第一张图"}, { "type": "image_url", "image_url": {"url": "examples/images/image_1.jpeg"}, }, {"type": "text", "text": "有什么内容?"}, ], }] prompt, media = encode_messages( messages, thinking_mode="chat", return_multi_modal_data=True, ) # prompt: # '<|begin▁of▁sentence|><|User|>第一张图\n\n<|deepseek_image|>\n\n有什么内容?<|Assistant|>' # media["images"] contains the image records in prompt order ``` Images are represented in the prompt by `<|deepseek_image|>`. `media["images"]` contains the corresponding image records in exactly the same order they appear in the prompt. Pixel loading and expansion into model image tokens are handled by `inference/image_processor.py`. ## Compact TXT notation `parse_tagged_text()` converts a compact prompt such as ```text 第一张图examples/images/image_1.jpeg有什么内容? ``` into the same standard content blocks. It is an input convenience layer, not a second encoding implementation. ## Message format ### Special tokens | Token | Purpose | | :--- | :--- | | `<|begin▁of▁sentence|>` | Beginning of sequence (BOS) | | `<|end▁of▁sentence|>` | End of assistant turn (EOS) | | `<|User|>` | User turn prefix | | `<|Assistant|>` | Assistant turn prefix | | `<|System|>` | Mid-conversation system message prefix | | `<|latest_reminder|>` | Latest reminder (date, locale, etc.) | | `` / `` | Reasoning block delimiters | | `|DSML|` | DSML markup token | | `<|deepseek_image|>` | Image placeholder in the prompt string | ### Roles The encoding supports the following message roles: `system`, `user`, `assistant`, `tool`, and `latest_reminder`. A `tool` message is not rendered directly: `merge_tool_messages()` converts it into a `` block inside the preceding user message. When multiple tool results are present, they are sorted by the order of the corresponding `tool_calls` in the preceding assistant message. ### Basic chat A simple multi-turn conversation is encoded as: ``` <|begin▁of▁sentence|>{system_prompt} <|User|>{user_message}<|Assistant|>{response}<|end▁of▁sentence|> <|User|>{user_message_2}<|Assistant|>{response_2}<|end▁of▁sentence|> ``` - The BOS token is prepended at the very beginning of the conversation. - In **chat mode** (`thinking_mode="chat"`), `` is placed right after `<|Assistant|>` to immediately close the thinking block, so the model generates content directly. ### Thinking mode In **thinking mode** (`thinking_mode="thinking"`), the model produces explicit reasoning inside `...` blocks before responding. ``` <|begin▁of▁sentence|><|System|>{reasoning_effort_prefix}{system_prompt} <|User|>{message}<|Assistant|>{reasoning}{response}<|end▁of▁sentence|> ``` The reasoning effort prefix is injected once, before the system message, as a `<|System|>` block: ``` <|System|>Reasoning Effort: {budget} (range 1-100, the higher the value, the more thorough the reasoning) ``` The `drop_thinking` parameter (default `True`) controls whether reasoning from earlier turns is preserved: - **Without tools**: reasoning content from assistant turns **before** the last user message is stripped. Only the final assistant turn retains its `...` block. - **With tools**: `drop_thinking` is automatically disabled. All turns retain their reasoning, because tool-calling conversations require full context for the model to track multi-step reasoning across tool calls. ### Tool calling (DSML format) Tools are defined on the `system` message via the `tools` field (OpenAI-compatible format). When tools are present, the following schema block is injected into the system prompt: ``` ## Tools You have access to a set of tools to help answer the user's question. You can invoke tools by writing a "<|DSML| calls>" block like the following: <|DSML| calls> <|DSML| invoke name="$TOOL_NAME"> <|DSML| parameter name="$PARAMETER_NAME" string="true|false">$PARAMETER_VALUE ... <|DSML| invoke name="$TOOL_NAME2"> ... String parameters should be specified as is and set `string="true"`. For all other types (numbers, booleans, arrays, objects), pass the value in JSON format and set `string="false"`. If thinking_mode is enabled (triggered by ), you MUST output your complete reasoning inside ... BEFORE any tool calls or final response. Otherwise, output directly after with tool calls or final response. ### Available Tool Schemas {tool_definitions_json} You MUST strictly follow the above defined tool name and parameter schemas to invoke tool calls. ``` An actual tool call in the assistant turn looks like: ```xml <|DSML| calls> <|DSML| invoke name="function_name"> <|DSML| parameter name="param" string="true">string_value <|DSML| parameter name="count" string="false">5 <|end▁of▁sentence|> ``` - `string="true"`: the parameter value is a raw string. - `string="false"`: the parameter value is JSON (number, boolean, array, object). Tool execution results are wrapped in `` tags within user messages: ``` <|User|>{result_json}<|Assistant|>... ``` ### Tool namespaces Tool definitions may include a `namespace` alongside `function`, either as a string or as an object with `name` and an optional `description`: ```python tool = { "type": "function", "namespace": {"name": "search", "description": "Search tools."}, "function": { "name": "lookup", "description": "Look up a value", "parameters": {"type": "object", "properties": {"query": {"type": "string"}}}, }, } tool_call = { "type": "function", "namespace": "search", "function": {"name": "lookup", "arguments": '{"query": "value"}'}, } ``` The tool schema and DSML invocation both use `search::lookup`. The namespace description is prepended to the tool description, separated by a newline. The parser returns `function.name="lookup"` and `namespace="search"` on the tool call, so its output can be passed back to `encode_messages()` directly. Input also accepts `namespace` inside `function`, or a qualified function name such as `search::lookup`. A qualified name must agree with any explicit namespace; `::` separates exactly one namespace from the tool name. Tools without a namespace retain their original names and output format. ### Reasoning effort Pass `reasoning_effort` as an integer in `[1, 100]` or as one of `"low"` (50), `"high"` (75), or `"max"` (100). The default is `"high"` (75). The setting only affects `thinking_mode="thinking"` and is only rendered at the start of the conversation (index 0). Intermediate values may be used to elicit interpolated reasoning behavior. ### Quick instruction special tokens Quick instruction tokens are used for auxiliary classification and generation tasks. They are appended to messages via the `"task"` field to trigger specialized model behavior for a single-token or short-form output. | Special Token | Description | Format | |:---|:---|:---| | `<|action|>` | Determines whether the user prompt requires a web search or can be answered directly. | `...<|User|>{prompt}<|Assistant|><|action|>` | | `<|title|>` | Generates a concise conversation title after the first assistant response. | `...<|Assistant|>{response}<|end▁of▁sentence|><|title|>` | | `<|query|>` | Generates search queries for the user prompt. | `...<|User|>{prompt}<|query|>` | | `<|authority|>` | Classifies the user prompt's demand for source authoritativeness. | `...<|User|>{prompt}<|authority|>` | | `<|domain|>` | Identifies the domain of the user prompt. | `...<|User|>{prompt}<|domain|>` | | `<|read_url|>` | Determines whether each URL in the user prompt should be fetched and read. | `...<|User|>{prompt}<|read_url|>` | Usage in message format: - **`action`** on a user message: the `<|action|>` token is placed after the assistant prefix and thinking token, triggering a routing decision (e.g., "Search" or "Answer"). - **Other tasks** (`query`, `authority`, `domain`, `read_url`) on a user message: the task token is appended directly after the user content. - **`title`** on an assistant message: the `<|title|>` token is appended after the assistant's EOS. The next assistant message provides the generated title. ## Tests From this directory: ```bash python -m pytest -q test_encoding.py ``` Test cases are stored as paired JSON input / TXT expected-output files under `tests/`. The tests cover multi-turn conversations, tool calling, thinking mode, numeric reasoning effort, mid-conversation system messages, and multimodal image ordering. They include a check that the TXT and JSON examples encode to the same prompt and preserve the same image ordering.