mirror of
https://github.com/NanmiCoder/cc-haha
synced 2026-07-18 13:23:33 +08:00
92b01565a3
Common desktop screenshots can be taller than the old 2000px local resize threshold while still fitting provider request limits. Align the local image dimension cap with the Claude Vision API rejection threshold so Read and desktop attachments pass through already-small screenshots, while oversized or unprocessable images still use the existing fallback paths. Constraint: Claude Vision API allows image dimensions up to 8000px and still has a 5MB base64 image payload limit Rejected: Keep 2000px as a client-side hard limit | it rejects common screenshots that providers can handle Confidence: high Scope-risk: narrow Related: #615 Related: #677 Tested: bun test tests/imageResizer.test.ts Tested: bun test tests/imageResizer.test.ts tests/mediaRecoveryAndEstimation.test.ts src/services/api/errors.test.ts src/server/__tests__/proxy-transform.test.ts Tested: bun test src/server/__tests__/conversation-attachments.test.ts tests/imageResizer.test.ts Tested: bun run check:server
95 lines
3.4 KiB
TypeScript
95 lines
3.4 KiB
TypeScript
/**
|
|
* Anthropic API Limits
|
|
*
|
|
* These constants define server-side limits enforced by the Anthropic API.
|
|
* Keep this file dependency-free to prevent circular imports.
|
|
*
|
|
* Last verified: 2026-06-01
|
|
* Source: Claude Vision API docs and api/api/config.py
|
|
*
|
|
* Future: See issue #13240 for dynamic limits fetching from server.
|
|
*/
|
|
|
|
// =============================================================================
|
|
// IMAGE LIMITS
|
|
// =============================================================================
|
|
|
|
/**
|
|
* Maximum base64-encoded image size (API enforced).
|
|
* The API rejects images where the base64 string length exceeds this value.
|
|
* Note: This is the base64 length, NOT raw bytes. Base64 increases size by ~33%.
|
|
*/
|
|
export const API_IMAGE_MAX_BASE64_SIZE = 5 * 1024 * 1024 // 5 MB
|
|
|
|
/**
|
|
* Target raw image size to stay under base64 limit after encoding.
|
|
* Base64 encoding increases size by 4/3, so we derive the max raw size:
|
|
* raw_size * 4/3 = base64_size → raw_size = base64_size * 3/4
|
|
*/
|
|
export const IMAGE_TARGET_RAW_SIZE = (API_IMAGE_MAX_BASE64_SIZE * 3) / 4 // 3.75 MB
|
|
|
|
/**
|
|
* API-enforced maximum dimensions for image inputs.
|
|
*
|
|
* Note: The API internally resizes images whose long edge is larger than
|
|
* 1568px, but that is handled server-side and doesn't cause errors. Keep this
|
|
* limit aligned with the API's rejection threshold so common tall screenshots
|
|
* are not locally resized or rejected just because they exceed 2000px.
|
|
*
|
|
* The API_IMAGE_MAX_BASE64_SIZE (5MB) is the actual hard limit that causes
|
|
* API errors for ordinary screenshots before dimensions usually matter.
|
|
*/
|
|
export const IMAGE_MAX_WIDTH = 8000
|
|
export const IMAGE_MAX_HEIGHT = 8000
|
|
|
|
// =============================================================================
|
|
// PDF LIMITS
|
|
// =============================================================================
|
|
|
|
/**
|
|
* Maximum raw PDF file size that fits within the API request limit after encoding.
|
|
* The API has a 32MB total request size limit. Base64 encoding increases size by
|
|
* ~33% (4/3), so 20MB raw → ~27MB base64, leaving room for conversation context.
|
|
*/
|
|
export const PDF_TARGET_RAW_SIZE = 20 * 1024 * 1024 // 20 MB
|
|
|
|
/**
|
|
* Maximum number of pages in a PDF accepted by the API.
|
|
*/
|
|
export const API_PDF_MAX_PAGES = 100
|
|
|
|
/**
|
|
* Size threshold above which PDFs are extracted into page images
|
|
* instead of being sent as base64 document blocks. This applies to
|
|
* first-party API only; non-first-party always uses extraction.
|
|
*/
|
|
export const PDF_EXTRACT_SIZE_THRESHOLD = 3 * 1024 * 1024 // 3 MB
|
|
|
|
/**
|
|
* Maximum PDF file size for the page extraction path. PDFs larger than
|
|
* this are rejected to avoid processing extremely large files.
|
|
*/
|
|
export const PDF_MAX_EXTRACT_SIZE = 100 * 1024 * 1024 // 100 MB
|
|
|
|
/**
|
|
* Max pages the Read tool will extract in a single call with the pages parameter.
|
|
*/
|
|
export const PDF_MAX_PAGES_PER_READ = 20
|
|
|
|
/**
|
|
* PDFs with more pages than this get the reference treatment on @ mention
|
|
* instead of being inlined into context.
|
|
*/
|
|
export const PDF_AT_MENTION_INLINE_THRESHOLD = 10
|
|
|
|
// =============================================================================
|
|
// MEDIA LIMITS
|
|
// =============================================================================
|
|
|
|
/**
|
|
* Maximum number of media items (images + PDFs) allowed per API request.
|
|
* The API rejects requests exceeding this limit with a confusing error.
|
|
* We validate client-side to provide a clear error message.
|
|
*/
|
|
export const API_MAX_MEDIA_PER_REQUEST = 100
|