返回博客
blog••
2

大模型 API 生态日报 —— 2026-11-19

无人值守每日汇总大模型 API 生态:4 条发布动态、2 个新模型、19 项价格变动与 1 条社区热议。

一、今日速览

  • 覆盖模型:479
  • 新增模型:2
  • 价格变动:19
  • 社区信号:1

本篇日报由 AINode 无人值守内容引擎自动生成,信源包括公开模型库、GitHub 官方发布与开发者社区讨论。

二、新增模型

模型 ID厂商上下文输入 /1M输出 /1M
google/gemini-4-neon google 1.0M $3.00 $15.00
google/gemini-4-neon:batch google 1.0M $1.50 $7.50

三、价格变动

  • moonshotai/kimi-k3.1-pro 输入 $0.240 → $0.253 (+5%)
  • moonshotai/kimi-k3.1-pro 输出 $8.40 → $8.80 (+5%)
  • deepseek/deepseek-v4.1-flash-1017 输入 $0.036 → $0.150 (+317%)
  • deepseek/deepseek-v4.1-flash-1017 输出 $0.500 → $0.600 (+20%)
  • ~deepseek/deepseek-flash-latest 输入 $0.036 → $0.035 (−3%)
  • ~deepseek/deepseek-flash-latest 输出 $0.500 → $0.150 (−70%)
  • ~deepseek/deepseek-v4-flash-latest 输入 $0.043 → $0.047 (+7%)
  • ~deepseek/deepseek-v4-flash-latest 输出 $0.300 → $0.140 (−53%)
  • deepseek/deepseek-v4-flash-0731 输入 $0.060 → $0.062 (+3%)
  • deepseek/deepseek-v4-flash-0731 输出 $0.180 → $0.300 (+67%)
  • z-ai/glm-5.2 输入 $0.0070 → $0.620 (+8757%)
  • z-ai/glm-5.2 输出 $6.80 → $2.42 (−64%)

四、官方与开源发布动态

  • vllm-project/vllm v0.34.0:v0.34.0
    • 10 new models (Mistral Large 4, MiniMax-M3.1-Flash, Nemotron-3.5-Transcribe, and more!)
    • Extensible KV cache is on by default: vLLM sizes the KV cache from memory actually free after warmup, so most models get a bigger cache and more throughput - in most cases you should now remove gpu_memory_utilization from your configs.
    • Model Runner V1 is gone. Model Runner V2 is now the only GPU runner.
    • Big speedups on Blackwell and MI355X: up to 1.95× decode (Step5 on GB200), 1.6× throughput (GLM-5.3-Flash), 1.24–1.40× faster long-prompt TTFT (GLM-5.3).
    查看发布 ↗
  • ollama/ollama v0.40.17:v0.40.17
    • Improve ChatGPT CLI restore and graceful restart by @hoyyeva in #19583
    • cmd/launch: use thinking metadata for OpenCode settings by @hoyyeva in #19409
    • llm: use mmap for Vulkan integrated GPUs by @dhiltgen in #19640
    • rpreslar4765/Ollama by @rpreslar4765 in #17163
    查看发布 ↗
  • deepseek-ai/DeepSeek-V3 v1.0.0:v1.0.0 查看发布 ↗
  • meta-llama/llama-models v0.2.0:v0.2.0 查看发布 ↗

五、社区热议

六、网关小贴士

统一 OpenAI 兼容入口与单一密钥,跨厂商做故障转移、Token 计量与按模型计费,无需逐个对接厂商 SDK。