返回博客
blog••
0

大模型 API 生态日报 —— 2026-11-21

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:476
  • 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
thinkingmachines/inkling-2-mini thinkingmachines 1.0M $0.100 $0.400
radixark/daisyrose-27b radixark 262K $0.150 $1.88

三、价格变动

  • nvidia/nemotron-3.5-super-vl 输出 $0.500 → $0.470 (−6%)
  • ~deepseek/deepseek-pro-latest 输入 $0.119 → $0.100 (−16%)
  • deepseek/deepseek-v4.1-flash 输入 $0.300 → $0.050 (−83%)
  • deepseek/deepseek-v4.1-flash 输出 $1.20 → $0.199 (−83%)
  • deepseek/deepseek-v4-flash-vision-exp 输入 $0.440 → $0.216 (−51%)
  • deepseek/deepseek-v4-flash-vision-exp 输出 $1.32 → $0.647 (−51%)
  • google/gemma-4-26b-a4b-it 输入 $0.090 → $0.068 (−25%)
  • google/gemma-4-26b-a4b-it 输出 $0.300 → $0.225 (−25%)
  • minimax/minimax-m2.7 输入 $0.210 → $0.600 (+186%)
  • minimax/minimax-m2.7 输出 $0.840 → $1.80 (+114%)
  • minimax/minimax-m2.5 输入 $0.295 → $0.270 (−8%)
  • minimax/minimax-m2.5 输出 $1.20 → $0.950 (−21%)

四、官方与开源发布动态

  • ollama/ollama v0.40.17:v0.40.17
    • ollama launch claude supports Claude Code's deferred tool loading, so MCP tools are loaded when needed instead of all at once
    • Ollama on Apple silicon now serves multiple requests at once.
    • ollama launch opencode sets thinking options from the model's metadata
    • ollama launch pi refreshes Pi's model list after sign-in and passes along prompt arguments
    查看发布 ↗
  • 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 ↗
  • 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。