Protocol projections | LLMChat Completions, Messages and Responses, each projected onto the neutral turn for a declared subset, in both directions, refusing what the subset does not carry.LLMexplanationllmexplanationdeveloperoperatorbuild-agent-systemsdeploy-operate-productsreference

Protocol projections

ExplanationLLM

ForDeveloper · Operator

ExperimentalPublic

Source & revision

llm/website/docs/concepts/protocols.md · 4af9aecb5786

A projection maps the neutral turn onto one wire protocol and back. Each of the three is a separate crate, and each serves both directions: an outgoing request to a model endpoint, and an incoming request from a client that speaks that wire. One codec for both directions means a client and a future gateway cannot disagree about the subset.

Crate	Protocol	Request path	Outgoing client	Ingress decoder
b10x-llm-chat	Chat Completions	{base}chat/completions	ChatClient	decode_ingress_request
b10x-llm-messages	Messages	{base}messages	MessagesClient	decode_request
b10x-llm-responses	Responses	{base}responses	none: a pure codec	ingest_request

ChatClient and MessagesClient implement Model: they resolve the credential for the attempt, send one request through the bounded transport and decode the stream. llm-responses performs no I/O at all. It builds request bodies (project_request) and decodes streamed payloads (decode_stream), and the caller supplies the HTTP exchange.

Rules all three follow

• Refuse, never drop. A request field outside the subset is refused by name. A translation that silently loses a field only looks like it succeeded.
• The binding decides authentication. anonymous sends no credential header, bearer sends Authorization: Bearer …, and api-key sends the header the account names. The protocol decides none of this, and none of it changes billing.
• The wire gets the upstream model name. The request carries the binding's upstream_name, never the caller's route alias.
• Absent stays absent. A sampling setting the caller did not set is not written, so the provider's default stays the provider's decision.
• Counters are normalized once. Each projection maps its wire's usage fields onto the neutral counters. A counter the wire did not report stays unknown. A zero the wire reported is kept as a report.
• Failures after dispatch say so. A refusal raised while reading a response never claims the request was not-sent, so a caller is never invited to resend a turn the provider may already have billed.

Chat Completions

The widest reach: any server that speaks this wire, including an anonymous local vLLM server.

• A streamed request always sends stream_options: {"include_usage": true}. Without it the wire reports no usage, which would look the same as a model that consumed nothing.
• max_output_tokens becomes max_completion_tokens.
• A failed tool result travels inside the text-only tool message as {"error": <output>, "ok": false}.
• delta.reasoning_content, which vLLM reasoning parsers emit, becomes a reasoning delta and is never joined to the answer text.
• Opaque continuation state is refused: this wire has nowhere to carry it.

Messages

• One route, POST {base}messages, in streaming mode, with a pinned anthropic-version: 2023-06-01 header.
• The wire requires max_tokens. If the caller sets no limit, the binding's declared maximum is sent, because that is the operator's statement and not a guess.
• A temperature above 1.0 is refused, because this wire's range is narrower than the neutral one.
• Signed and redacted thinking cross as opaque state bound to all six coordinates. On ingress they arrive unattributed and are not sendable until the caller binds them.
• The stream has no [DONE] sentinel. message_stop ends it, and a stream that simply stops is refused with its last valid snapshot attached.

Responses

• Requests are always streamed (stream: true) and never stored (store: false). The caller owns the conversation and replays it whole each turn.
• The standing instruction becomes a leading developer message.
• Tool results are always enveloped as {"ok": <bool>, "output": <value>}, successes included, so a round trip cannot flip the failed flag.
• A turn ends only at a response.completed or response.incomplete payload. The end of the stream is not a terminal answer.
• Every request asks for reasoning.encrypted_content. Reasoning entries a client sends in are carried unattributed, like Messages thinking.
• OpenAI streams reasoning as response.reasoning_summary_* events and vLLM as response.reasoning_*. Both spellings become reasoning deltas.
• An unrecognized stream event is kept as an opaque item and reported as a warning, not dropped.

What this evidence covers

Each projection has its own conformance suite and a recorded set of deliberate mutations that the suite catches. All of it runs against pinned response bytes, loopback sockets and in-process fakes. None of it has been run against a live provider. The subset each crate implements is listed in full in its repository document: chat, messages and responses.

Call a local endpoint runs ChatClient against a local server.