* fix(relay): set Request.GetBody so the HTTP/2 transport can transparently retry after an upstream stream reset
The outbound request body is a type-erased io.Reader over BodyStorage, so
net/http cannot derive Request.GetBody (it only does so for *bytes.Reader,
*bytes.Buffer and *strings.Reader). With GetBody nil, the HTTP/2 transport
cannot transparently retry a request once the body has been written and the
upstream resets the stream with a retryable error (REFUSED_STREAM, or a
connection-level GOAWAY); the relay request then fails with:
http2: Transport: cannot retry err [...] after Request.Body was written;
define Request.GetBody to avoid this error
This affects every relay path that goes through DoApiRequest (chat, claude,
gemini, responses, embedding, image, rerank).
BodyStorage (memory and disk) already implements io.Seeker, so replay support
only needed wiring:
- NewOutboundJSONBody additionally returns a getBody that rewinds the storage
and hands out a fresh non-closing reader. The transport only calls GetBody
after the previous attempt's body has been abandoned, so the rewind cannot
race an in-flight read.
- RelayInfo carries it in the new UpstreamRequestGetBody field, set alongside
UpstreamRequestBodySize by the handlers that build storage-backed bodies.
- applyUpstreamGetBody (symmetric with applyUpstreamContentLength) wires it
into DoApiRequest/DoFormRequest/DoTaskApiRequest, only when req.GetBody is
still nil.
Also remove the hand-rolled GetBody override in DoTaskApiRequest: it returned
the same already-consumed reader, so any transport-level replay would have
silently sent an empty body, and it clobbered the correct snapshot-based
GetBody that net/http derives from the *bytes.Reader bodies the task adaptors
pass in. For non-replayable bodies GetBody now stays nil, so a retry fails
loudly instead of corrupting the request.
Covered by unit tests plus an end-to-end raw-frame HTTP/2 test that resets
the first stream with REFUSED_STREAM after the body is written and asserts
the transport transparently retries with the complete body.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(relay): hand out independent readers from GetBody (address review)
Per the http.Request.GetBody contract ("returns a new copy of Body"),
each call must yield a reader with its own cursor. The previous
implementation rewound and reused the shared BodyStorage, so two
consecutive GetBody readers would interfere with each other, and a
replay could disturb the primary body's offset under extreme transport
timing (e.g. attempt N's body write not yet fully abandoned when the
transport builds attempt N+1).
Instead of snapshotting the payload (an extra copy), add
BodyStorage.NewReader, which returns an independent zero-copy reader:
- memory mode: a fresh bytes.Reader over the same immutable backing
array;
- disk mode: a separate file descriptor over the cache file, so the
transport closing a replayed body only closes that descriptor.
NewOutboundJSONBody's getBody now simply hands out storage.NewReader,
and once the handler releases the storage, GetBody fails with
ErrStorageClosed instead of replaying stale data.
Tests: interleaved reads across two replay readers and the primary
body each observe exactly their own byte stream, for both the memory
and the disk-backed storage; the existing GetBody and HTTP/2 retry
suites still pass (h2 e2e tests flake-free with -count=20).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(relay): bind replayable metadata on pass-through requests
* fix(relay): reset upstream body metadata between channels
* test(relay): cover replay across retries and channel attempts
* fix(relay): stop following upstream redirects
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
* test(relayconvert): add golden snapshot matrix and relaykit boundary guard
Phase 0 of the relaykit extraction plan: pin byte-level output of every
registered (from,to) request/response/stream conversion route, and
forbid kit-bound packages from growing host-only imports.
* wip(relayconvert): drop gin.Context from converter signatures; add convmeta draft
Phase 1 in progress: relayconvert now takes context.Context; host media
resolver adapts gin.Context back at the service boundary.
* refactor(relayconvert): decouple converters from RelayInfo, gin, and settings
Phase 1 of the relaykit extraction plan:
- converters now depend on convmeta.Meta (implemented by RelayInfo) instead
of *relaycommon.RelayInfo; ClaudeConvertInfo and the format guesser move
to convmeta with aliases left behind
- host settings reach converters via a convmeta.Options snapshot built in
RelayInfo.ConvOptions; no more model_setting/reasoning global reads inside
the conversion layer
- effort-suffix helpers move to service/relayconvert/reasoning (old package
forwards); chat-to-responses upgrade policy moves to service (host routing
logic, not conversion)
- golden conversion matrix unchanged
* test(relayconvert): tighten boundary — kit packages now free of gin/setting imports
* refactor(dto): drop gin and logger dependencies
Phase 2 (part 1): dto.Request.IsStream now takes *http.Request instead of
*gin.Context (Gemini's impl reads query/path off the std request); dto's
three logger calls become common.SysError. Boundary test allowlist is now
empty — kit-bound packages import no gin/setting/logger/model.
* refactor(kit): extract dependency-free kitutil; dto/types/relayconvert stop importing common
Phase 2 of the relaykit extraction plan:
- new service/relayconvert/kitutil holds the pure helpers the kit needs
(JSON wrappers, pointer/string/uuid/timestamp utils, MaskSensitiveInfo,
pluggable LogInfo/LogError hooks, Debug flag)
- dto, types, and all relayconvert packages now use kitutil; their only
remaining internal deps are dto/types/constant
- common keeps every original symbol (MaskSensitiveInfo delegates to
kitutil) so host code is untouched; main.go routes kit logging into
common.SysLog/SysError and mirrors DebugEnabled
- golden conversion matrix unchanged
* refactor(kit): move EndpointType/FinishReason to types; OpenRouter dialect via Options
Kit packages (dto/types/relayconvert/reasonmap) no longer import constant:
- EndpointType and finish-reason values live in types; constant re-exports
- the OpenRouter special-case in claude->openai request conversion reads
Options.OpenRouterDialect, set by the host from the channel type;
InitChannelMeta invalidates the cached snapshot on channel switch
* refactor: extract relaykit submodule (dto/types/relayconvert/reasonmap)
Phase 3 of the relaykit extraction plan:
- new go module github.com/QuantumNous/new-api/relaykit containing dto
(minus task family), types, relayconvert (with convmeta/kitutil/reasoning),
and reasonmap; host consumes it via require + replace, go.work for dev
- task-family dto (task/suno/midjourney/video) stays in the host dto
package; dual-consumer host files alias it as taskdto
- relaykit builds and tests standalone (GOWORK=off): no host imports,
no gin, no DB, no settings
- golden conversion matrix unchanged
* build(docker): copy relaykit/go.mod before go mod download
The local-replace submodule's go.mod must exist inside the build context
for the main module graph to resolve.
* fix: address relaykit extraction regressions
* fix: address relaykit review regressions
* docs: document Meta nil receiver contract
* fix(relaykit): fail OpenAI→Claude conversion without max_tokens; reject negative default_max_tokens
The Claude Messages API requires max_tokens (omitting it is a 400
"Field required"), but with a nil Options.Claude.DefaultMaxTokens hook
the converters silently emitted a request the upstream is guaranteed to
reject. Both OpenAI Chat and Responses → Claude conversions now return
sharedclaude.ErrMissingMaxTokens when no path (client value, default
hook, thinking-adapter floor) supplied one. Unreachable in the host,
which always configures the hook.
Host side, claude.default_max_tokens now rejects negative values at the
option API before persisting — they would wrap into huge unsigned values
during conversion. Zero stays allowed: the current API treats
max_tokens: 0 as cache pre-warming.
* fix: make Gemini safety settings read path race-free
Bound max-tokens fields across all relay format validators, saturate
tiered-expression rounding and audio/tool/task token conversions, and
route legacy remix ratios through the guarded setter.
Bound user-supplied count/duration parameters at request validation,
route ratio multipliers through guarded setters, and use saturating
int conversions in all quota math paths.
- Added `EstimateBilling` function to check for video input in request metadata and return corresponding discount ratios.
- Updated `ModelPriceHelperPerCall` to incorporate new pricing logic based on model ratios and video input.
- Enhanced task billing logs to include model ratio information and adjusted calculations for actual quota based on additional multipliers.
- Introduced `renderTaskBillingProcess` to improve rendering of task billing information in the UI.
- Refactored the request URL and body construction methods to align with the Veo predictLongRunning endpoint.
- Introduced new data structures for Veo instances and parameters, replacing the previous Gemini video generation configurations.
- Updated the Vertex adaptor to utilize the new Veo request payload format.
- Added Gemini video generation configuration structures and payloads.
- Introduced functions for parsing and resolving video duration and resolution from metadata.
- Enhanced the Vertex adaptor to support Gemini video generation requests and billing estimation based on duration and resolution.
- Updated model pricing settings for new Gemini video models.
1. Async task model redirection (aligned with sync tasks):
- Integrate ModelMappedHelper in RelayTaskSubmit after model name
determination, populating OriginModelName / UpstreamModelName on RelayInfo.
- All task adaptors now send UpstreamModelName to upstream providers:
- Gemini & Vertex: BuildRequestURL uses UpstreamModelName.
- Doubao & Ali: BuildRequestBody conditionally overwrites body.Model.
- Vidu, Kling, Hailuo, Jimeng: convertToRequestPayload accepts RelayInfo
and unconditionally uses info.UpstreamModelName.
- Sora: BuildRequestBody parses JSON and multipart bodies to replace
the "model" field with UpstreamModelName.
- Frontend log visibility: LogTaskConsumption and taskBillingOther now
emit is_model_mapped / upstream_model_name in the "other" JSON field.
- Billing safety: RecalculateTaskQuotaByTokens reads model name from
BillingContext.OriginModelName (via taskModelName) instead of
task.Data["model"], preventing billing leaks from upstream model names.
2. Per-call billing (TaskPricePatches lifecycle):
- Rename TaskBillingContext.ModelName → OriginModelName; add PerCallBilling
bool field, populated from TaskPricePatches at submission time.
- settleTaskBillingOnComplete short-circuits when PerCallBilling is true,
skipping both adaptor adjustments and token-based recalculation.
- Remove ModelName from TaskSubmitResult; use relayInfo.OriginModelName
consistently in controller/relay.go for billing context and logging.
3. Multipart retry boundary mismatch fix:
- Root cause: after Sora (or OpenAI audio) rebuilds a multipart body with a
new boundary and overwrites c.Request.Header["Content-Type"], subsequent
calls to ParseMultipartFormReusable on retry would parse the cached
original body with the wrong boundary, causing "NextPart: EOF".
- Fix: ParseMultipartFormReusable now caches the original Content-Type in
gin context key "_original_multipart_ct" on first call and reuses it for
all subsequent parses, making multipart parsing retry-safe globally.
- Sora adaptor reverted to the standard pattern (direct header set/get),
which is now safe thanks to the root fix.
4. Tests:
- task_billing_test.go: update makeTask to use OriginModelName; add
PerCallBilling settlement tests (skip adaptor adjust, skip token recalc);
add non-per-call adaptor adjustment test with refund verification.
- Renamed RelayTask function to RelayTaskFetch for clarity.
- Updated routing in relay-router.go and video-router.go to use RelayTaskFetch for fetch operations.
- Enhanced error handling in RelayTaskFetch function.
- Adjusted task data conversion in TaskAdaptor to include task ID.
Add three billing lifecycle methods to the TaskAdaptor interface:
- EstimateBilling: compute OtherRatios from user request before pricing
- AdjustBillingOnSubmit: adjust ratios from upstream submit response
- AdjustBillingOnComplete: determine final quota at task terminal state
Introduce BaseBilling as embeddable no-op default for adaptors without
custom billing. Move Sora/Ali OtherRatios logic from shared validation
into per-adaptor EstimateBilling implementations.
Add TaskBillingContext to persist pricing params (model_price, group_ratio,
other_ratios) in task private data for async polling settlement.
Extract RecalculateTaskQuota as a general-purpose delta settlement
function and unify polling billing via settleTaskBillingOnComplete
(adaptor-first, then token-based fallback).
Restructure the task relay system for better separation of concerns:
- Extract task billing into service/task_billing.go with unified settlement flow
- Move task polling loop from controller to service/task_polling.go (supports Suno + video platforms)
- Split RelayTask into fetch/submit paths with dedicated retry logic (taskSubmitWithRetry)
- Add TaskDto, TaskResponse generics, and FetchReq to dto/task.go
- Add taskcommon/helpers.go for shared task adaptor utilities
- Remove controller/task_video.go (logic consolidated into service layer)
- Update all task adaptors (ali, doubao, gemini, hailuo, jimeng, kling, sora, suno, vertex, vidu)
- Simplify frontend task logs to use new TaskDto response format
Clamp request body size (including post-decompression) to avoid memory exhaustion caused by huge payloads/zip bombs, especially with large-context Claude requests. Add a configurable `MAX_REQUEST_BODY_MB` (default `32`) and document it.
- Enforce max request body size after gzip/br decompression via `http.MaxBytesReader`
- Add a secondary size guard in `common.GetRequestBody` and cache-safe handling
- Return **413 Request Entity Too Large** on oversized bodies in relay entry
- Avoid building large `TokenCountMeta.CombineText` when both token counting and sensitive check are disabled (use lightweight meta for pricing)
- Update READMEs (CN/EN/FR/JA) with `MAX_REQUEST_BODY_MB`
- Fix a handful of vet/formatting issues encountered during the change
- `go test ./...` passes
CaIon