jay77721 dc6c0e5de5 fix: drop max_tokens from generic OpenAI-compatible request builder (#17857) ...

## Summary

The generic `buildRequestBody` in `internal/entity/models/base_model.go`
unconditionally forwarded `ChatConfig.MaxTokens` as `"max_tokens"` for
every OpenAI-compatible provider.

Providers that need a different token field already delete or override
it after the call (e.g. Xiaomi uses `max_completion_tokens`, Replicate
uses `max_new_tokens`). This change stops setting `max_tokens` in the
shared builder so it only forwards the parameters common across
providers, and each provider remains free to set its own token limit
field.

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

2026-08-05 19:31:12 +08:00

..

302ai_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

302ai.go

Go: fix warnings (#17738)

2026-08-03 21:30:01 +08:00

aliyun_test.go

fix: implement Tongyi-Qianwen TTS via DashScope OpenAI-compatible endpoint (#17770)

2026-08-04 16:39:31 +08:00

aliyun.go

fix: implement Tongyi-Qianwen TTS via DashScope OpenAI-compatible endpoint (#17770)

2026-08-04 16:39:31 +08:00

anthropic_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

anthropic.go

Go: introduce content_length and max_output (#17807)

2026-08-04 19:08:31 +08:00

astraflow_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

astraflow.go

Go: fix warnings (#17738)

2026-08-03 21:30:01 +08:00

avian_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

avian.go

feat(go-models): migrate batch 1 model drivers to unified handlers (#17696)

2026-08-03 12:13:43 +08:00

azure_openai_test.go

refactor(go-models): migrate remaining OpenAI-compatible drivers to shared HTTP pipeline (#17787)

2026-08-04 16:59:25 +08:00

azure_openai.go

refactor(go-models): migrate remaining OpenAI-compatible drivers to shared HTTP pipeline (#17787)

2026-08-04 16:59:25 +08:00

baichuan.go

Go: fix warnings (#17738)

2026-08-03 21:30:01 +08:00

baidu.go

Go: fix warnings (#17738)

2026-08-03 21:30:01 +08:00

base_model.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

bedrock_test.go

feat(go-models): migrate batch 6 model drivers to unified usage recording (#17775)

2026-08-04 14:07:45 +08:00

bedrock.go

feat(go-models): migrate batch 6 model drivers to unified usage recording (#17775)

2026-08-04 14:07:45 +08:00

builtin.go

Go: fix warnings (#17738)

2026-08-03 21:30:01 +08:00

chat_tools.go

Go: add context to lots of interface (#17253)

2026-07-22 22:30:57 +08:00

chat_usage_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

cohere_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

cohere.go

Go: fix warnings (#17738)

2026-08-03 21:30:01 +08:00

cometapi_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

cometapi.go

feat(go-models): migrate batch 1 model drivers to unified handlers (#17696)

2026-08-03 12:13:43 +08:00

common_test.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

common.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

deepinfra_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

deepinfra.go

feat(go-models): migrate batch 1 model drivers to unified handlers (#17696)

2026-08-03 12:13:43 +08:00

deepseek_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

deepseek.go

refactor(go-models): migrate remaining OpenAI-compatible drivers to shared HTTP pipeline (#17787)

2026-08-04 16:59:25 +08:00

dummy.go

Go: add context to lots of interface (#17253)

2026-07-22 22:30:57 +08:00

factory.go

feat: add GreenPT model provider (#17447)

2026-07-28 19:19:00 +08:00

fishaudio_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

fishaudio.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

funasr_test.go

fix(provider): support editing fetched model types before save (#17728)

2026-08-03 18:07:41 +08:00

funasr.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

futurmix.go

feat(go-models): migrate batch 1 model drivers to unified handlers (#17696)

2026-08-03 12:13:43 +08:00

get_driver_test.go

fix: docx/email parsing, extractor LLM driver, and chunker alignment (#17144)

2026-07-21 13:51:17 +08:00

get_driver.go

fix: docx/email parsing, extractor LLM driver, and chunker alignment (#17144)

2026-07-21 13:51:17 +08:00

gitee_test.go

feat: update all_models.json to content_length + max_output (#17839)

2026-08-05 15:43:11 +08:00

gitee.go

feat(go-models): migrate batch 2 model drivers to unified handlers (#17697)

2026-08-03 13:54:35 +08:00

google_test.go

feat(go-models): migrate batch 6 model drivers to unified usage recording (#17775)

2026-08-04 14:07:45 +08:00

google.go

feat(go-models): migrate batch 6 model drivers to unified usage recording (#17775)

2026-08-04 14:07:45 +08:00

gpustack_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

gpustack.go

refactor(go-models): migrate remaining OpenAI-compatible drivers to shared HTTP pipeline (#17787)

2026-08-04 16:59:25 +08:00

greenpt_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

greenpt.go

feat(go-models): migrate batch 2 model drivers to unified handlers (#17697)

2026-08-03 13:54:35 +08:00

groq_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

groq.go

refactor(go-models): migrate remaining OpenAI-compatible drivers to shared HTTP pipeline (#17787)

2026-08-04 16:59:25 +08:00

huaweicloud_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

huaweicloud.go

Go: fix warnings (#17738)

2026-08-03 21:30:01 +08:00

huggingface.go

feat(go-models): migrate batch 2 model drivers to unified handlers (#17697)

2026-08-03 13:54:35 +08:00

hunyuan_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

hunyuan.go

feat(go-models): migrate batch 2 model drivers to unified handlers (#17697)

2026-08-03 13:54:35 +08:00

jiekouai_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

jiekouai.go

feat(go-models): migrate batch 2 model drivers to unified handlers (#17697)

2026-08-03 13:54:35 +08:00

jina_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

jina.go

fix: jina’s incorrect return value: modelname (#17716)

2026-08-03 18:07:55 +08:00

llm_test.go

fix(go-agent): message double (#17353)

2026-07-24 19:23:51 +08:00

llm.go

fix(go-agent): message double (#17353)

2026-07-24 19:23:51 +08:00

lmstudio.go

feat(go-models): migrate batch 2 model drivers to unified handlers (#17697)

2026-08-03 13:54:35 +08:00

localai_test.go

feat(go-models): migrate batch 2 model drivers to unified handlers (#17697)

2026-08-03 13:54:35 +08:00

localai.go

feat(go-models): migrate batch 2 model drivers to unified handlers (#17697)

2026-08-03 13:54:35 +08:00

longcat_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

longcat.go

refactor(go-models): migrate remaining OpenAI-compatible drivers to shared HTTP pipeline (#17787)

2026-08-04 16:59:25 +08:00

mineru_local.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

mineru.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

minimax_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

minimax.go

refactor(go-models): migrate remaining OpenAI-compatible drivers to shared HTTP pipeline (#17787)

2026-08-04 16:59:25 +08:00

mistral_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

mistral.go

feat(go-models): migrate batch 3 model drivers to unified handlers (#17698)

2026-08-03 15:08:55 +08:00

model_test_helpers_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

model_test.go

feat: populate content_length and max_output from official docs (#17836)

2026-08-05 15:22:57 +08:00

model.go

Go: introduce content_length and max_output (#17807)

2026-08-04 19:08:31 +08:00

modelscope_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

modelscope.go

feat(go-models): migrate batch 3 model drivers to unified handlers (#17698)

2026-08-03 15:08:55 +08:00

moonshot_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

moonshot.go

refactor(go-models): migrate remaining OpenAI-compatible drivers to shared HTTP pipeline (#17787)

2026-08-04 16:59:25 +08:00

n1n_test.go

fix(go-models): harden N1N default transport handling (#15351)

2026-06-02 13:40:10 +08:00

n1n.go

feat(go-models): migrate batch 3 model drivers to unified handlers (#17698)

2026-08-03 15:08:55 +08:00

non_chat_usage_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

novita_test.go

fix(go-models): record Novita streaming usage without chatConfig (#17778)

2026-08-04 14:42:16 +08:00

novita.go

fix(go-models): record Novita streaming usage without chatConfig (#17778)

2026-08-04 14:42:16 +08:00

nvidia_rerank_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

nvidia_test.go

Go: introduce content_length and max_output (#17807)

2026-08-04 19:08:31 +08:00

nvidia.go

Go: introduce content_length and max_output (#17807)

2026-08-04 19:08:31 +08:00

ollama_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

ollama.go

feat(go-models): migrate batch 3 model drivers to unified handlers (#17698)

2026-08-03 15:08:55 +08:00

openai_api_compatible.go

feat(go-models): migrate batch 3 model drivers to unified handlers (#17698)

2026-08-03 15:08:55 +08:00

openai_compatible_usage_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

openai_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

openai.go

refactor(go-models): migrate remaining OpenAI-compatible drivers to shared HTTP pipeline (#17787)

2026-08-04 16:59:25 +08:00

openrouter_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

openrouter.go

feat(go-models): migrate batch 3 model drivers to unified handlers (#17698)

2026-08-03 15:08:55 +08:00

orcarouter.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

paddleocr_local.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

paddleocr.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

parser_config.go

feat(go-models): migrate batch 3 model drivers to unified handlers (#17698)

2026-08-03 15:08:55 +08:00

perplexity_test.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

perplexity.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

ppio_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

ppio.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

provider_response_usage_test.go

feat(go-models): migrate batch 6 model drivers to unified usage recording (#17775)

2026-08-04 14:07:45 +08:00

qiniu_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

qiniu.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

ragcon_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

ragcon.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

reasoning_family_provider_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

replicate_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

replicate.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

response_handler_test.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

response_handler.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

siliconflow_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

siliconflow.go

refactor(go-models): migrate remaining OpenAI-compatible drivers to shared HTTP pipeline (#17787)

2026-08-04 16:59:25 +08:00

sse_scanner_buffer_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

stepfun_test.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

stepfun.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

timeout_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

togetherai_test.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

togetherai.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

tokenhub_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

tokenhub.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

tokenpony_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

tokenpony.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

tool_call_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

types.go

Go: introduce content_length and max_output (#17807)

2026-08-04 19:08:31 +08:00

upstage_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

upstage.go

feat(go-models): migrate batch 4 model drivers to unified handlers (#17699)

2026-08-03 17:48:45 +08:00

usage_parser_test.go

feat(go-models): unify OpenAI-compatible token usage extraction (#17634)

2026-07-31 18:32:07 +08:00

usage_parser.go

feat(go-models): unify OpenAI-compatible token usage extraction (#17634)

2026-07-31 18:32:07 +08:00

vllm_rerank_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

vllm.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

volcengine_test.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

volcengine.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

voyage_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

voyage.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

xai_test.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

xai.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

xiaomi_test.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

xiaomi.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

xinference_test.go

fix: drop max_tokens from generic OpenAI-compatible request builder (#17857)

2026-08-05 19:31:12 +08:00

xinference.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

xunfei_test.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

xunfei.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00

zhipu_ai_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

zhipu-ai_test.go

Go: fix SSRF (#17641)

2026-07-31 19:15:38 +08:00

zhipu-ai.go

feat(go-models): migrate batch 5 model drivers to unified handlers (#17700)

2026-08-03 20:19:22 +08:00