ext/new-api
2
0
Fork 0
mirror of https://github.com/QuantumNous/new-api.git synced 2026-09-11 14:41:21 +00:00
Code Activity

feat(relay): hosted-tool conversion fidelity, reasoning normalization, and billing usage integrity (#7137)

Browse Source 
* feat(relaykit): preserve hosted tools across conversions

- add protocol-neutral hosted-tool DTOs, conversion metadata, and loss policies
- bridge citations, grounding metadata, and hosted-tool stream lifecycles
- document the public conversion behavior and channel policy controls

* refactor(relaykit): normalize reasoning and thinking intent

- centralize provider-neutral reasoning intent, effort, and budget mappings
- parse model suffixes at the host entry boundary while preserving provider-owned tails
- keep adaptive Claude thinking and explicit zero-token compatibility consistent

* fix(billing): preserve authoritative usage across relay hops

- carry native BillingUsage sidecars through direct and streamed protocol bridges
- merge partial and terminal usage monotonically with safe fallback settlement
- retain cache metadata, penultimate usage, and per-call Gemini tool surcharges

* feat(relay): bridge Responses with Claude and Gemini protocols

- add direct request, response, and stream converters across supported relay formats
- expose Claude count_tokens and Chat-to-Responses compatibility endpoints
- carry conversion diagnostics through the host while retaining the curated public goldens

* fix(relay): wire relaykit conversions into host channels

- connect handlers, adaptors, and channel settings to the standalone conversion layer
- keep model mapping, pricing identity, retries, and provider-specific suffix behavior aligned
- ignore local audit artifacts and retain focused public regression coverage
This commit is contained in:
Calcium-Ion
2026-09-01 21:53:35 +08:00
committed by GitHub
parent b7017c251b
commit 0ed497f066
149 changed files with 14729 additions and 3044 deletions
Download Patch File Download Diff File Expand all files Collapse all files
setting/reasoning/suffix.go
+14 -6
View File
@@ -1,9 +1,12 @@
// Package reasoning re-exports the pure model-name effort-suffix helpers,
// which moved to the conversion kit (service/relayconvert/reasoning) as part
// which moved to the conversion kit (relaykit/relayconvert/reasoning) as part
// of the relaykit extraction. Host code keeps importing this path unchanged.
package reasoning
import kitreasoning "github.com/QuantumNous/new-api/relaykit/relayconvert/reasoning"
import (
kitreasoning "github.com/QuantumNous/new-api/relaykit/relayconvert/reasoning"
"github.com/QuantumNous/new-api/setting/model_setting"
)
var (
EffortSuffixes = kitreasoning.EffortSuffixes
@@ -12,8 +15,13 @@ var (
)
var (
TrimEffortSuffix = kitreasoning.TrimEffortSuffix
TrimEffortSuffixWithSuffixes = kitreasoning.TrimEffortSuffixWithSuffixes
ParseOpenAIReasoningEffortFromModelSuffix = kitreasoning.ParseOpenAIReasoningEffortFromModelSuffix
ParseDeepSeekV4ThinkingSuffix = kitreasoning.ParseDeepSeekV4ThinkingSuffix
TrimEffortSuffixWithSuffixes = kitreasoning.TrimEffortSuffixWithSuffixes
ParseDeepSeekV4ThinkingSuffix = kitreasoning.ParseDeepSeekV4ThinkingSuffix
TrimGeminiThinkingSuffix = kitreasoning.TrimGeminiThinkingSuffix
)
// ParseOpenAIReasoningEffortFromModelSuffix applies the host effort-tail
// whitelist so real model IDs such as qwen-max are not treated as aliases.
func ParseOpenAIReasoningEffortFromModelSuffix(modelName string) (string, string) {
return kitreasoning.ParseOpenAIReasoningEffortFromModelSuffix(modelName, model_setting.ShouldPreserveEffortTail)
}