Skip to content

BFE v1.8.6

Latest
Latest

Choose a tag to compare

View all tags
@cc14514 cc14514 released this 31 Aug 10:23
v1.8.6
9dce1ac
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

What's New in v1.8.6

Added

  • Support Anthropic Claude protocol forwarding with model protocol matching (#1331)
  • Support RMB tiered pricing with peak/off-peak tiers (#1335)
  • Support cache token billing for AI requests (#1330)
  • Support audio token RMB billing (#1331)
  • Support image generation per-image billing (#1331)
  • Recognize DeepSeek cache fields in usage parsing (#1337)
  • Add session-level API-Key affinity within the same cluster (#1337)
  • Add req_body_larger_than and req_body_less_than condition primitives (#1336)
  • Add GetInt64Batch to redis_client for batch quota remaining reads (#1337)
  • Use control-plane generated redis_key for AI rate limit counters (#1331)
  • Support full Redis key cleanup exported by rate-limit export (#1333)
  • Add BFE AI gateway error codes documentation (#1339)

Changed

  • Change session-level API-Key affinity TTL to idle timeout with default 600s (#1337)
  • Simplify token_rule.data fields and add TagLevel in mod_ai_token_auth (#1329)

Fixed

  • Fix model modification when retry between clusters in ServeHTTPForAI (#1334)
  • Fix PriceMap/TierPriceMap JSON serialization to preserve 8-decimal precision (#1338)

Full Changelog: https://github.com/bfenetworks/bfe/blob/v1.8.6/CHANGELOG.md

Assets 6
Loading