Three layered optimizations targeting Gemini-style 5MB base64 payloads where
RSS could balloon to tens of GB under concurrent load:
1. Byte-based param override (relay/common/override.go)
- Switch legacy/operations hot paths from common.Marshal round-trips and
map[string]any conversions to gjson/sjson on []byte directly.
- Avoids cloning 5MB strings during each Set/Delete operation.
2. strings.Builder for Gemini response markdown (relay/channel/gemini/relay-gemini.go)
- Replace string concatenation + strings.Join when assembling
"" content for inline image responses.
- Pre-allocates capacity from inline_data byte sizes.
3. Outbound BodyStorage + streaming Decoder (this commit's core)
- New relay/common/outbound_body.go helper wraps marshaled upstream bodies
in common.BodyStorage, allowing disk-cache mode to offload jsonData to
a temp file while waiting for upstream TTFB. The original []byte can
then be GC'd, removing ~5MB/req of heap residency during the longest
window of a request.
- All 7 relay handlers (gemini/claude/responses/embedding/image/compatible/
rerank) plus chat_completions_via_responses adopt the helper with
defer closer.Close() and explicit jsonData = nil.
- relay/common/relay_info.go: new UpstreamRequestBodySize so
relay/channel/api_request.go can populate req.ContentLength (lost when
body becomes a type-erased io.Reader).
- common/gin.go UnmarshalBodyReusable: when storage is disk-backed and
content-type is JSON, decode via DecodeJson(storage) instead of
storage.Bytes()+Unmarshal, removing one transient 5MB copy per request.
memory mode and form/multipart paths unchanged.
When proxying through another new-api instance, the upstream
X-Oneapi-Request-Id was overwriting the local one in client responses.
This adds a new `upstream_request_id` field to the logs table, captures
the upstream ID during relay, and filters it from being copied back to
the client. Frontend gains search/filter and detail display support.
The Azure channel's GetRequestURL method only handled RelayModeResponses
but missed RelayModeResponsesCompact. This caused compact requests to
fall through to the generic deployments URL pattern, producing an
incorrect path that Azure returns 404 for.
This fix extends the existing responses API special handling to also
cover the compact mode, appending /compact to the subUrl when the relay
mode is ResponsesCompact.
Affected URLs (before → after):
- Normal Azure: /openai/deployments/{model}/responses/compact → /openai/v1/responses/compact
- cognitiveservices: same pattern → /openai/responses/compact
- Custom AzureResponsesVersion: properly respected for compact too
Co-authored-by: 彭俊杰 <pengjunjie@onero.com>
- 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.
* feat: add HEIC/HEIF image format support
Add detection, MIME type mapping, and dimension parsing for HEIC/HEIF
images via ISOBMFF ftyp brand inspection and ispe box parsing. Update
Gemini relay to accept these formats and refactor getImageConfig to
properly retry decoders using buffered data.
* fix: handle ISOBMFF extended sizes in HEIF dimension parser
parseHEIFDimensions now correctly handles boxSize==1 (64-bit extended
size) and boxSize==0 (box-to-EOF), preventing the parser from breaking
out of the loop when encountering these valid ISOBMFF box headers
before reaching the meta box.
Add detection, MIME type mapping, and dimension parsing for HEIC/HEIF
images via ISOBMFF ftyp brand inspection and ispe box parsing. Update
Gemini relay to accept these formats and refactor getImageConfig to
properly retry decoders using buffered data.
The previous commit commented out AddOtherRatio("n") in the Ali
adaptor to fix double-counting but this could cause billing evasion
when n is specified via extra["parameters"] instead of request.N.
Root cause: ImagePriceRatio in GetTokenCountMeta() already included
n, AND channel adaptors added OtherRatio("n"), resulting in n²
billing.
Proper fix:
- Remove n from ImagePriceRatio (keep sizeRatio * qualityRatio only)
- In ImageHelper, add default OtherRatio("n") when adaptor hasn't
set one; set fallback tokens to 1 (base unit)
- Restore Ali adaptor's AddOtherRatio("n") — it uses actual upstream
parameters/response count, preventing billing evasion
- Add StreamStatus type (relay/common) to track stream end reason
(done/timeout/client_gone/scanner_error/eof/panic/ping_fail) and
accumulate soft errors during streaming via sync.Once + sync.Mutex.
- Add StreamResult (relay/helper) as the callback interface: adapters
call sr.Error() for soft errors, sr.Stop() for fatal, sr.Done() for
normal completion. No early-return problem — multiple errors per chunk
are naturally supported.
- Refactor StreamScannerHandler callback from func(string) bool to
func(string, *StreamResult). All 9 channel adapters updated.
- Write stream_status into log other JSON field (admin-only) with
status ok/error, end_reason, error_count, and error messages.
- Frontend: display stream status in log detail expansion for admins.
