feat: add kling video o1 support - #2491
Conversation
…d size validation
feat: EditTokenModal 中针对用户创建的 token 默认无限额度
feat: add environment variable switch for critical rate limit
…80p-image fix: trim suffix p for jimeng image model
…space fix: tag splitting by whitespace
…odisable fix(channel): 当没有可用密钥时返回错误而不是第一个密钥
… new sections for partners, acknowledgments, and deployment instructions
- 更新中文README.md中的语言链接 - 完全重写英文README.en.md，包含所有详细功能说明 - 完全重写法文README.fr.md，确保内容一致性 - 完全重写日文README.ja.md，提供完整的项目说明 所有语言版本现在具有： - 相同的结构和格式 - 一致的语言导航 - 完整的功能特性和部署指南 - 统一的环境变量配置说明
…ence 修复viduq2不支持参考生视频的问题
…annel feat: replicate channel flux model
fix GetChannelKey AdminAuth -> RootAuth
…-tts feat: support gpt tts series model quota calculate
…in GetAudioDuration" This reverts commit e293be0.
…itelist-cidr feat(auth): enhance IP restriction handling with CIDR support
… meta when token count is disabled 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
Tighten oversized request handling across relay paths and make error matching reliable. - Align `MAX_REQUEST_BODY_MB` fallback to `32` in request body reader and decompression middleware - Stop ignoring `GetRequestBody` errors in relay retry paths; return consistent **413** on oversized bodies (400 for other read errors) - Add `Unwrap()` to `types.NewAPIError` so `errors.Is/As` can match wrapped underlying errors - `go test ./...` passes
…-response-id fix: 模型设置增加针对Vertex渠道过滤content[].part[].functionResponse.id的选项，默认启用
- Replace legacy `docs.newapi.pro` paths with the new `/{lang}/docs/...` structure across all README translations
- Point key sections (installation, env vars, API, support, features) to their new locations
- Ensure language-specific links use the correct locale prefix (zh/en/ja) and keep FR aligned with English routes
Keep new-site links (/{lang}/docs/...) where matching pages exist in the current docs repo
Revert links that have no equivalent in the new docs to the legacy paths on doc.newapi.pro:
Google Gemini Chat
Midjourney-Proxy image docs
Suno music docs
Apply the same rule consistently across all README translations (zh/en/ja/fr)
…te-doc-links-new-routing 🔗 docs(readme): update documentation links to new site routing
WalkthroughTwo files modified to extend Kling adaptor functionality: added three new data structures (ImageItem, ElementItem, VideoItem) and extended requestPayload with corresponding lists; modified BuildRequestURL to select Omni-Video path for "o1" model names; changed BuildRequestBody to obtain requests via relaycommon.GetTaskRequest; altered relay_utils unmarshalling to unconditionally invoke UnmarshalBodyReusable after multipart handling. Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes
Possibly related PRs
Suggested reviewers
Poem
Pre-merge checks and finishing touches❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✨ Finishing touches
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (1)
relay/channel/task/kling/adaptor.go (1)
56-70: LGTM! Well-documented Omni-Video structures.The new data structures are clearly defined with helpful inline comments documenting the expected values for enum-like fields (
type,refer_type,keep_original_sound).Optional: Consider typed constants for enum fields
To improve type safety and prevent invalid values, you could define constants for the enum-like fields:
// Frame type constants const ( FrameTypeFirst = "first_frame" FrameTypeEnd = "end_frame" ) // Refer type constants const ( ReferTypeFeature = "feature" ReferTypeBase = "base" ) // Sound preservation options const ( KeepSoundYes = "yes" KeepSoundNo = "no" )However, this is purely optional—string fields are acceptable for external API bindings.
📜 Review details
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
relay/channel/task/kling/adaptor.gorelay/common/relay_utils.go
🧰 Additional context used
🧬 Code graph analysis (1)
relay/common/relay_utils.go (1)
common/gin.go (1)
UnmarshalBodyReusable(72-97)
🔇 Additional comments (2)
relay/channel/task/kling/adaptor.go (2)
88-91: LGTM! Proper field definitions for Omni-Video support.The new fields are correctly defined with:
- Appropriate
omitemptytags to avoid sending empty arrays- Clear naming that matches the Omni-Video API contract
- Proper integration with the existing requestPayload structure
171-174: Use GetTaskRequest for cleaner request retrieval.Using
relaycommon.GetTaskRequestis a good refactor that improves clarity and error handling. The call ordering is guaranteed by the orchestration logic inrelay_task.go, which always invokesValidateRequestAndSetActionbeforeBuildRequestBody, ensuring the request is stored in context before retrieval.
| var path string | ||
| if strings.Contains(info.OriginModelName, "o1") { | ||
| path = "/v1/videos/omni-video" | ||
| } else { | ||
| path = lo.Ternary(info.Action == constant.TaskActionGenerate, "/v1/videos/image2video", "/v1/videos/text2video") | ||
| } |
There was a problem hiding this comment.
Strengthen model name matching to avoid false positives.
Using strings.Contains(info.OriginModelName, "o1") to detect Omni-Video models could match unintended model names. For example:
- Future models like
kling-pro1orkling-v101would incorrectly match - Models containing "o1" in other contexts would be misrouted
Consider using a more specific check:
🔎 Proposed fix using suffix or exact matching
Option 1: Check for known o1 model names
var path string
-if strings.Contains(info.OriginModelName, "o1") {
+if strings.HasSuffix(info.OriginModelName, "-o1") || info.OriginModelName == "kling-video-o1" {
path = "/v1/videos/omni-video"
} else {
path = lo.Ternary(info.Action == constant.TaskActionGenerate, "/v1/videos/image2video", "/v1/videos/text2video")
}Option 2: Use a helper function for clarity
+func isOmniVideoModel(modelName string) bool {
+ return strings.HasSuffix(modelName, "-o1")
+}
+
var path string
-if strings.Contains(info.OriginModelName, "o1") {
+if isOmniVideoModel(info.OriginModelName) {
path = "/v1/videos/omni-video"
} else {
path = lo.Ternary(info.Action == constant.TaskActionGenerate, "/v1/videos/image2video", "/v1/videos/text2video")
}🤖 Prompt for AI Agents
In relay/channel/task/kling/adaptor.go around lines 141 to 146, replace the
brittle strings.Contains(info.OriginModelName, "o1") check with a precise
matcher: either test against a set of known Omni-Video model names (map or slice
of exact names), or use suffix/exact matching or a regex anchored to avoid
partial matches (e.g., endsWith or ^...$). Encapsulate this logic in a small
helper like isOmniVideoModel(name string) bool and use it here so only intended
model names (or suffixes) route to "/v1/videos/omni-video"; otherwise choose the
image2video/text2video paths as before.
| } | ||
| // 为了metadata字段的兼容性，统一UnmarshalBodyReusable | ||
| if err := common.UnmarshalBodyReusable(c, &req); err != nil { | ||
| return createTaskError(err, "invalid_request", http.StatusBadRequest, true) | ||
| } |
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
# Find and examine relay/common/relay_utils.go around lines 209-217
if [ -f "relay/common/relay_utils.go" ]; then
echo "=== relay/common/relay_utils.go (lines 200-225) ==="
sed -n '200,225p' relay/common/relay_utils.go
fi
# Find validateMultipartTaskRequest
echo ""
echo "=== Searching for validateMultipartTaskRequest ==="
rg "validateMultipartTaskRequest|ValidateBasicTaskRequest" --type=go -C 3
# Find parseMultipartFormData
echo ""
echo "=== Searching for parseMultipartFormData ==="
rg "parseMultipartFormData" --type=go -C 10
# Find UnmarshalBodyReusable
echo ""
echo "=== Searching for UnmarshalBodyReusable ==="
rg "UnmarshalBodyReusable" --type=go -C 10Repository: QuantumNous/new-api
Length of output: 40743
🏁 Script executed:
#!/bin/bash
# Get more details on parseMultipartFormData implementation
sed -n '440,520p' common/gin.go
# Also check validateMultipartTaskRequest implementation
rg "func validateMultipartTaskRequest" -A 40 --type=goRepository: QuantumNous/new-api
Length of output: 2365
🏁 Script executed:
#!/bin/bash
# Get the parseMultipartFormData implementation from common/gin.go
rg "func parseMultipartFormData" -A 50 --type=go common/gin.goRepository: QuantumNous/new-api
Length of output: 1455
🏁 Script executed:
#!/bin/bash
# Check if there are any data: URL handling patterns in this area
rg "data:" relay/common/relay_utils.go -B 2 -A 2
# Check the task_video.go file mentioned in learnings
if [ -f "controller/task_video.go" ]; then
echo "=== Checking controller/task_video.go for data: URL pattern ==="
rg "data:" controller/task_video.go -B 2 -A 2 | head -40
fi
# Check what happens to the req after both parses
sed -n '205,235p' relay/common/relay_utils.goRepository: QuantumNous/new-api
Length of output: 1130
🏁 Script executed:
#!/bin/bash
# Check if processFormMap could cause different behavior
rg "func processFormMap" -A 30 --type=go common/gin.go
# Check the request body handling to understand reusability
rg "GetRequestBody\|KeyRequestBody" -B 2 -A 5 --type=go common/gin.go | head -50Repository: QuantumNous/new-api
Length of output: 588
🏁 Script executed:
#!/bin/bash
# Verify the complete flow by checking what happens to the req struct
# Check if req from validateMultipartTaskRequest is actually used before UnmarshalBodyReusable overwrites it
sed -n '200,225p' relay/common/relay_utils.go
# Also check if there are any tests showing expected behavior
rg "ValidateBasicTaskRequest" -A 5 --type=go *test*.go 2>/dev/null | head -30Repository: QuantumNous/new-api
Length of output: 885
🏁 Script executed:
#!/bin/bash
# Final verification: check if there are any side effects from validateMultipartTaskRequest besides the returned req
# that might justify the double-parsing
rg "validateMultipartTaskRequest" -B 5 -A 15 relay/common/relay_utils.go
# Check if storeTaskRequest uses the req before UnmarshalBodyReusable is called
rg "storeTaskRequest" -B 2 -A 10 relay/common/relay_utils.go | head -30Repository: QuantumNous/new-api
Length of output: 1861
Confirm the need for sequential multipart parsing and eliminate redundancy.
For multipart requests, the code parses the form data twice:
- First via
validateMultipartTaskRequest(line 209), which callsc.MultipartForm() - Then via
UnmarshalBodyReusable(line 215), which re-parses from raw body bytes via customparseMultipartFormData
The first parse's result is completely overwritten by the second, creating unnecessary performance overhead. Either use the first parse result directly, or remove the first parse if the second serves the metadata compatibility requirement mentioned in the comment.
20 participants
新增可灵 kling-v2-6和kling-video-o1视频模型支持
请求示例:
Summary by CodeRabbit
Release Notes
New Features
Improvements
✏️ Tip: You can customize this high-level summary in your review settings.